A dominant focus of our work is lighting for pedestrians and cyclists – using road lighting after dark to make walking and cycling safer and more attractive, whilst being respectful of the need to reduce the broader impact of lighting after dark.
Recent news items show public concern with road safety, such as poor car parking which hinders travel by wheelchair. A similar issue for cyclists is vehicle parking on cycle lanes, so it is good to see plans which aim to counter this behaviour. A recent article on road lighting highlights a general misunderstanding of what we should be able to expect lighting to do, or not do. Or, perhaps, it highlights that road lighting is a convenient target for complaining about other problems in one’s life.
Our project SATURN (Supporting Active Travel Using Road-lighting at Night) seeks to show the degree to which road lighting of different illuminances enhances the numbers of cyclists, and these data will be fed into the traffic planning model developed by Robin Lovelace at Leeds University. Jim Uttley gave a talk on the SATURN project at the Cycle City Active City conference in Sheffield on 6th July 2022. The slides from this talk are available here.
In May we had a visit from Kynthia Chamilothori of Eindhoven Technical University. Our PhD students gave presentations about their work, following which Kynthia gave feedback on what they might do to improve their work. Our two newest PhD students, Shahab Gorjimahlabani and Maan Balela recently completed their confirmation reviews and so are now on the path towards thesis submission.
EPSRC granted our project HAROLD (HAzards, ROad Lighting and Driving) a 12-month extension in lieu of COVID. This is good news; it means we have another year to work with Yichong.
The past few months and those ahead have been busy for our PhD students. Three are currently writing up their theses, with submission deadlines in late summer: Scott Fox’s work on driving and hazards, Intisar Hussain’s work on windows and privacy, and Khalid Hamoodh’s work on interpersonal evaluations.
Four students are mid-way through their work. Shahab Gorjimahlabani has just finished a further reassurance field study. He is testing Boyce’s day-dark method by defining day and dark periods at the same time of day by taking advantage of daylight-savings clock changes, rather than day and dark being different times of day.
Maan Balela has been able to collate travel count data for cyclists from a wide range of locations. He is using these data to test some of the assumptions of the analysis as well as looking to explain variance between locations. Aysheh Alshdaifat and Nima Hafezparast Moadab, the early career researchers on the LightCAP project, finished their first experiment testing the influence of lighting on measures of alertness in the late evening.
All of them have been affected by travel restrictions, building closure and workplace restrictions imposed following the Covid-19 pandemic, and all have done well to cope with the additional stress of these limitations.
Chloe Robbins, researcher on the HAROLD project, left to join the Home Office. She was replaced by Yichong Mao who had recently completed his thesis on similar work. Yichong’s thesis explored multi-tasking of typical pedestrian tasks such as hazard detection while HAROLD has a focus on the degree to which distraction hinders hazard detection by drivers.
HAROLD started just before the onset of the pandemic and was therefore substantively affected by restrictions on laboratory based experimental work. Yichong has conducted a couple of pilot studies and we are collaborating with Natasha Merat and colleagues at Leeds University to conduct experiments in parallel using their driving simulator.
As pandemic restrictions around the world begin to be lifted we can look ahead to attending once again in-person conferences. Over the next few months we will be visiting IAPS, ICTTP7, Lux Europa and the annual meeting of CIE Division 4. For PhD students intending to join LumeNet 2022 in September, please register soon so that we can plan the review panel.
In recent years our work has focussed mainly on outdoor lighting to support active travel – walking and cycling. One aspect of this research is pedestrian reassurance – using road lighting to enhance the confidence to walk after dark. We have recently produced a short film (below) to explain this work and our methods of research.
You might also be interested in listening to a discussion on pedestrian reassurance between Steve Fotios, Allan Howard and Chris Fordham of The Light Review.
In October we held a presentation session for the PhD students – the first for a couple of years. Scott Fox started his PhD with the aim of investigating the influence of lighting and distraction on hazard detection when driving. This was set back by building closure during the coronavirus pandemic, which prevent access to the laboratory, and he is now focussing more directly on the influence of distraction to hazard detection.
Khalid Hamoodh is investigating the visual cues used when evaluating the intent other pedestrians. While past studies have assumed this was the face, this assumption has not been previously tested. In other words, Khalid is conducting research to inform how we should do research about lighting. Both Scott and Khalid are working now towards submitting their theses.
The focus of Maan Belela’s work is cycling. His first question is the influence of ambient light level on cycling rates, and testing the variation between different locations. Shahab Gorjimahlabani is investigating pedestrian reassurance. In the first place, he is using a range of methods to better understand how these influence the conclusions drawn. His first experiment followed the travel count approach but used on-road observation rather than automated counters so that we could also capture the age and gender of travellers.
A second experiment followed the qualitative approach: discussion of why some locations do or do not feel safe to walk after dark. At the current time, his current experiment is using the day-dark approach to measuring reassurance, introduced by Boyce et al., but extending this by questioning whether the day and dark periods are established.
Aysheh Alshdaifat and Nima Hafezparast Moadab, early career researchers on the LightCAP project, are investigating the influence of light on alertness after dark for pedestrians and drivers. If lighting can be used to enhance alertness, this could contribute to a reduction in frequency and severity of road traffic crashes.
The PhD research methods symposium this year is the Academic Forum which focuses on daylight research. It is an online event on 18 November. Please contact Jens Christoffersen to sign up.
In 2022, LumeNet will be held (hopefully) in Sheffield as an in-person event. We are lining up an excellent team of expert reviewers, so far including Peter Boyce, Kevin Houser and John O’Hagan. LumeNet typically takes place over two days and covers a wide range of topics in lighting. For 2022 we are planning an additional day, supported by CIE-UK, for presentations to a wider audience.
Research is of limited value when it is read only by other researchers. To have impact, research also needs to reach the people it is intended to benefit. We show here some recent examples.
“One of the documents that I quite often refer to is [Road lighting research for drivers and pedestrians: The basis of luminance and illuminance recommendations by Steve Fotios and Ron Gibbons]... In my opinion [they are] probably two of the most influential people in road lighting in the world... They both regularly write papers, they regularly present... and both are fantastic presenters. It is well worth reading... ”
This suggests that the research has reached at least one lighting engineer and that he respects that research and it may then inform his future work. Any change to infrastructure can elicit a negative reaction from the general public, perhaps especially so for something as visible as road lighting, and that may limit the extent to which a designer makes changes from current provision. Informing the wider public of research may help to mitigate negative reactions.
One topic to which we have devoted much attention is pedestrian reassurance – how we can use lighting to enhance the feeling of safety when walking after dark. This might boost confidence to walk rather than to drive, or to walk rather than stay at home. This was the focus of a BBC report which included brief excerpts from an interview with Steve (the report starts at about 10.50 minutes).
Students at the University of Sheffield have started a campaign, Our Bodies Our Streets, focussing on safety concerns of female pedestrians after dark. The University of Sheffield journalism students reported the issue in The Star, a local newspaper, and their research included interviews with Steve and Jim.
Sage, the publisher of Lighting Research and Technology, commissioned a blog for Sage Perspectives to highlight a collection of papers on pedestrian lighting and perceptions of safety. While Sage Perspectives tends to target academics rather than the general public, it should reach a wider audience of academics than those who already tend to read LR&T.
Good news! We have recently been awarded funding from the Engineering and Physical Sciences Research Council (EPSRC) to carry out a new three-year project about the effects of road lighting on cycling rates and cyclist safety.
SATURN (Safer Active Travel Using Road-lighting at Night, research grant number EP/V043587/1) will gather data about cyclist numbers, collisions involving cyclists, and road lighting characteristics in a number of locations across the UK. The total value of the project is £0.6 million, of which £0.4 million is awarded to Sheffield and £0.2 million to our collaborators at the University of Leeds.
Cycling is good for the cyclist and is good for society in general. People who cycle rather than drive improve their physical fitness and general health. Cycling rather than driving reduces air pollution, reduces carbon emissions and reduces traffic congestion. It also benefits local economies, as research shows cycling-friendly neighbourhoods encourage more use of local shops.
Savings to public health costs through increased physical activity also means cycling benefits the national economy. The UK Government has recognised the benefits of cycling, particularly in response to the Covid-19 pandemic, through increased investment in cycling infrastructure and a new strategy to promote cycling.
Darkness puts some people off cycling, particularly women. For those who do cycle at night, there is a greater risk of being involved in a crash. Road lighting can mitigate both of these problems. Road lighting can improve the visibility of cyclists to drivers and thus reduce the frequency and severity of road traffic collisions. Road lighting encourages people to cycle when it is dark – they feel safer.
However, we do not know what lighting conditions best meet the needs of cyclists. Existing evidence about lighting for cyclists is weak, and current guidance about lighting for cyclists is based on out-of-date standards that focus on the needs of pedestrians rather than cyclists.
SATURN will provide evidence to help find the best lighting conditions for cycling, to both encourage people to cycle after-dark and to make them safer.
We will first measure the effect of darkness (compared with daylight) on the number of people cycling and the number of vehicle collisions involving a cyclist. This will be done using a method of analysis that isolates the effect of lighting from other influential factors, a method we have refined in recent pilot studies.
We will use cyclist count and collision data for a range of locations in six major cities. These data will come from a number of sources including automated cyclist counters, crowdsourced cycling trip data and the national database of road collisions. This will tell us the size of the effect darkness has on cycling rates and collision risk at each location.
We will then compare this to metrics of the lighting at each of the locations, measured using a mobile array of sensors and backed up using existing night-time aerial photography images. This comparison will reveal the relationship between lighting, cycling rates and safety, allowing us to identify the most appropriate lighting levels to support cycling.
Results from this analysis of cycling rates, collision rates and lighting will be added to an online planning tool designed to help local transport planners develop and prioritise decisions about cycling infrastructure in their local area. This will help ensure SATURN has real impact and influences transport planning decisions related to cycling infrastructure. We will also use the SATURN findings to influence revisions to national and international road lighting standards, to ensure these standards take account of the needs of cyclists using the best available evidence.
This has of course being a difficult year for laboratory-based research. Within the HAROLD project, Chris had improved the scale-model road scene with which to investigate the effect of distraction on target detection.
We were ready to start recruiting participants in March... and have not yet been able to continue this work. Luckily, the preliminary study of distraction using abstract targets was completed and this work has now been published.
Our review of distractions to driving suggested that conversation with passengers was the most prevalent distraction, according to previous studies reporting road-side observation or post-crash interviews. Comparison of this conclusion with a new study suggests that road type may be a significant factor; while passenger conversation is the most prevalent distraction on urban roads this may not be the case for trunk routes. Whilst not able to conduct trials in the lab, Chloe has instead carried out analyses using road traffic crash data and travel flow data – for example, see this investigation of motorcycle crashes.
Choong Yew Chang has just submitted his PhD thesis, a study of window view quality, and will soon attend his viva with Barabara Matusiak of the Norwegian University of Science and Technology. Aleks Liachenko Monteiro and Yichong Mao are finalising their theses and should also be submitting soon.
In January 2021 we will be joined by two new PhD students, Maan Balela who will be working on active travel, Aysheh Alshdaifat who will be investigating pedestrian crossings, and Nima Hafezparast Moadab who will investigate drivers and hazard detection. Aysheh and Nima are members of the LightCAP project. Looking forward to meeting them, and appreciate the effort they are making to move to a new country at this difficult time.
Promoting reproducible science
Imagine conducting an experiment which reaches interesting conclusions and then publishing a report of that work. One test of the validity of such conclusions is that the same conclusions are again revealed when the experiment is repeated. This is known as reproducible science.
If the repeated experiment does not reach the same conclusions as the original work, despite using an apparently identical design and levels of independent variables, then one or other of the experiments must be erroneous.
Conclusions which cannot be reproduced in further work may be an indication of insufficient (small) samples, HARKing or p-hacking in the original work, not unexpected given the pressure on researchers to produce new and original findings with impact.
Robust research methodology is a key focus of work in the Lighting Research Group at the University of Sheffield. We have shown, for example, how range bias in an adjustment task explains the outcome of brightness preference experiments, colour preference and measurement of discomfort from glare.
We have repeated experiments on multiple occasions to determine whether the findings are robust. Consider investigation of pedestrians’ ability to detect pavement hazards. Our original experiment was repeated four times, checking an extended range of illuminances, an alternative mode of presenting hazards, the type (raised or depressed) of trip hazard (Fotios et al. 2020), and a recently completed study investigating the impact of multi-tasking.
We have published several papers promoting reproducible lighting research and the July 2020 Editor’s report for LEUKOS, The Journal of the Illuminating Engineering Society of North America, suggests that these are having some impact.
In the year to that report, the most downloaded paper was Jim Uttley’s article ‘Power analysis, sample size, and assessment of statistical assumptions: Improving the evidential value of lighting research’. This article reviewed previously published lighting research and found that most studies use inadequate sample sizes and fail to report important statistical information such as how statistical assumptions are tested, how sample sizes are determined, and the size of any effects found. The article also provided a brief primer on some of these issues, to help improve statistical reporting within the field in the future.
In that same period, the most cited paper was Steve Fotios’ article ‘A revised Kruithof graph based on empirical data’. The original Kruithof graph proposed a relationship between the colour (CCT) and amount (illuminance) of light for comfortable conditions. It has since been widely cited in design texts and has been apparently ‘validated’ by many other researchers.
However, even a cursory read of Kruithof’s original paper would reveal the absence of anything like robust evidence – there are no clues as to what experiment he did. In this new paper, Steve searched for papers directly or indirectly investigating the Kruithof graph and screened them using a very basic set of methodological criteria, for example, that in a repeated measures design the different scenes were observed in a randomised order and that conclusions were established through statistical analysis.
Only nine of the 29 papers considered passed this screening process. When the results from these papers were plotted, the resultant graph was a flat line, a very different graph to that of Kruithof, and one which suggests there is no relationship.
One problem with repeating a previous experiment is that by using their findings as a datum for the selection of test variables, the original findings may be repeated by default, leading to a misplaced conclusion of validated findings. Simply reproducing the results of a previous study does not always indicate that study provides good evidence, if it was based on flawed methodology. In the Lighting Research Group we focus on using robust methods that provide both valid and reproducible findings.
Supporting walking and cycling after dark
During the period of restricted mobility and social interaction more people are walking and cycling. The UK government appears to be taking advantage of this and is supporting infrastructure improvements to support walking and cycling after travel restrictions are lifted.
Encouraging people to walk or cycle rather than to drive, or not to leave the house at all, has major benefits for the whole population, including reduce air pollution and congestion (fewer motorised vehicles), improved personal health through physical activity, and reduced health expenditure.
Due to the seasonal change in times of sunrise and sunset, some journeys may be made after dark. With one response to Covid-19 being a proposal to stagger the working day this may lead to greater proportion of commuting taking place in hours of darkness.
Research over the past decade from the University of Sheffield’s Lighting Research Group has targeted lighting for walking and cycling after dark. Good lighting gives people the reassurance to walk or cycle – they think it will be safe. We have demonstrated this through qualitative research (stated preferences) and by counting travel flows under different lighting conditions (revealed preferences).
Good lighting can make it safer to walk or cycle – they are better able to see potential hazards. We have demonstrated this by testing the effect of changes in lighting on the key tasks of pedestrians and cyclists, as established using mobile eye tracking. We have used these findings to revise the guidance and technical reports issued by national and international bodies for lighting.
The findings of this work are described in our publications – many of which are open access so they can be read and downloaded without cost. Some of the key articles for walking and cycling after dark are:
Discussion of design guidance – limitations and proposals
Light encourages active travel
Uttley J, Fotios S, Lovelace R. Road lighting density and brightness linked with increased cycling rates after dark. PLOS One. In press. Accepted 29 April 2020 (preprint available).
Light makes us feel safer
Fotios S, Liachenko Monteiro A, Uttley J. Evaluation of pedestrian reassurance gained by higher illuminances in residential streets using the day-dark approach. Lighting Research and Technology, 2019; 51(4): 557-575.
Light can make it safer to walk or cycle
Fotios S, Johansson M. Appraising the intention of other people: Ecological validity and procedures for investigating effects of lighting for pedestrians. Lighting Research and Technology 2019; 51(1): 111-130.
Work continues through the virus lockdown! The main focus of our research is laboratory based experiments of visual reactions under changes in lighting: that is on hold until mobility restrictions are lifted. In the meantime we are pleased to announce that Hussain Qasem passed his PhD viva with only minor corrections, after examination by Eva Heinen and Richard Rowe. Hussain investigated lighting for cycling; amongst other things he used eye tracking to study where cyclists look and how different forms of lighting (road lighting and cycle lighting) affect the detection of hazards in peripheral vision.
A new interview was posted on BrightLights. Jennifer Veitch of the National Research Council Canada was interviewed by Juliëtte van Duijnhoven, who has recently completed her PhD and is now a postdoc at Eindhoven University of Technology. It is a brilliant interview, thanks to Juliëtte for good preparation.
LumeNet 2020 is scheduled to take place in Eindhoven in November 2020. Current planning is assuming that we will be able to travel by then. If you are a PhD student of lighting, colour, non-visual responses to light, or daylighting, I would suggest you register as soon as possible. Numbers are limited (to ensure sufficient time to focus on each student) and there may be competition for places from new researchers in the LightCAP project.
Steve has become the editor in chief of Lighting Research and Technology, following the departure of former EiC, Peter Boyce. As the most prolific author in LR&T over the past decade, Steve was an ideal candidate for the job. For research of lighting design, the human factors of lighting and lighting technology, LR&T (along with Leukos) is the key journal.
In December 2019 the CIE published a new technical report: CIE 236:2019. Lighting for pedestrians: A summary of empirical data. A large amount of the data in that report was the outcome of the MERLIN and MERLIN-2 projects at Sheffield (EPSRC grants EP/H050817 and EP/M02900X/1).
The report suggests that we can use light levels at the lower end of the range currently specified in the P-class. It provides evidence for local authorities, designers, and the authors of technical guidance, to counter the tendency for ever increasing light levels, and that is a benefit for energy use and the natural environment. To continue the work behind CIE 236:2019 we are holding a pedestrian lighting symposium in Sheffield, in September 2020.
Three of our PhD students (Hussain Qasem, Yichong Mao and Aleks Liachenko Monteiro) are now working towards submitting their PhD theses. These have concerned lighting for pedestrians and cyclists and contribute to our inputs to technical reports such as CIE 236:2019.
We established LumeNet to help PhD students at Sheffield and elsewhere through critical review from established experts in lighting research. The next event, LumeNet2020, will be held in Eindhoven, alongside Experiencing Light, and is being organised by Kynthia Chamilothori of Eindhoven Technical University.
In September we attended the International Conference on Environmental Psychology (ICEP) held at Plymouth University. Intisar Hussain presented a poster about the rating scale she developed to measure the clothing required for modesty/privacy in different contexts. This next leads to a discussion of the trade-off between clothing level and window blind closure and hence to the potential for exposure to daylight inside the home.
Scale for measuring the clothing privacy: each interval presents a step change in the degree of skin exposed or the tightness of clothing.
Jim organised a workshop on active travel, which included presentations from Khalid Hamoodh about interpersonal evaluations, Ian Walker of Bath University on the hidden needs and suppressed demand for active travel, and Peter Walley from Transport for Greater Manchester and Leeds University talking about communication between cyclists and drivers.
Elsewhere, Steve contributed a discussion on range bias to the sensory networks session and Chloe discussed her PhD work on drivers’ visual attention in a transport session.
Khalid, Scott, Jim, Chloe and Steve at ICEP2019
On Tuesday 8 October, the Daylight Academic Forum was held at Sorbonne University. Along with LumeNet, this is the ninth annual event we have held. The event gives PhD students an opportunity to discuss their research with world-leading experts in the field to gain constructive feedback. It was particularly good to see Kynthia Chamilothori return as an expert after attending previous events as a student. Kynthia has recently been awarded her doctorate and has moved to Technical University Eindhoven as an Associate Professor. Thanks once again to Jens Christoffersen for organising the Academic Forum.
Experts and students at the 2019 Daylight Academic Forum in Paris.
Steve is a member of the LightCAP consortium, the recently announced EU-funded training network. LightCAP focuses on the non-visual aspects of lighting, from fundamental research to application, and is led by Yvonne de Kort at Technical University Eindhoven. Yvonne and colleagues will be organising the next LumeNet (November 2020) alongside Experiencing Light.
In June we attended the 29th session of the CIE in Washington DC. Huge thanks to CIE-UK who supported our travel expenses, including the initiation of travel bursaries for PhD students.
Steve and Jim gave presentations about lighting to support pedestrians and cyclists. In the TC sessions Steve passed on news that the TC 4-52 report (Lighting for pedestrians) has been finalised and is now awaiting comments and approval from the board of directors. Jim established a reportership on lighting for cyclists (a reporter is often the precursor to a technical committee). Working through the CIE we are able to use our research findings to influence design standards.
Aleksandra, Yichong and Khalid presented posters of their PhD work; thanks to Professor Joanne Wood for questioning the students. Discussion is an important part of research, and an essential part of research training; LumeNet and the Velux Academic Forum were established to meet this need, providing critical review of research plans. Following the CIE conference the students presented their work at a CIE-UK meeting. CIE-UK are keen to support new researchers and have proposed to host a research presentation event for PhD students.
Chloe Robbins, RA for the HAROLD project, has submitted her PhD thesis at Nottingham University (in less than three years, an excellent achievement!) and from August will be working at Sheffield full time. Scott Fox, who will be working with Chloe to conduct experiments of detection under distraction for his PhD, recently attended an internship at TRL. He conducted a literature review to compare visual detection with automated detection.
Following the CIE Expert tutorial and workshop on research methods for human factors in lighting (Copenhagen, 2018), LEUKOS is publishing a special issue of papers on lighting research methodology. The Sheffield Lighting Research Group will have three papers in that special issue.
Veitch JA, Fotios SA, Houser KW. Judging the scientific quality of applied lighting research. Leukos: Special issue on research methods. Online first.
A discussion of research methods, whether planning new work or reviewing published work by others. This paper looks at research quality considerations, research design, internal and external validity, defining lighting conditions, statistical analysis and research ethics.
Fotios S. Using category rating to evaluate the lit environment: Is a meaningful opinion captured? Leukos: Special issue on research methods. Online first.
This discussion of category rating focuses on the assumed truth of responses. This assumption is that respondents have valid experience on which to base their response, have a valid recollection of that experience, have properly read the question, understand the question in the same way as the experimenter intended, and are then able to accurately indicate their opinion. In many cases, if not all, these are assumptions too far.
Uttley J (2019). Power analysis, sample size, and assessment of statistical assumptions: Improving the evidential value of lighting research. Leukos: Special issue on research methods.
Interpretation of experimental results should use statistical tests to determine whether differences and relationships are real and not just a chance effect. In this paper Jim Uttley describes some of the limitations common in lighting research including sample size analysis, testing the assumption of normality and reporting effects size.
Staff changes: Jim Uttley has returned to take up a lectureship in the School of Architecture, a position created by Karim Hadjri, the Head of School, to enable the Lighting Research Group to grow. Since the end of his previous RA post on the MERLIN project Jim has been working with Natasha Merat at Leeds University Institute for transport studies. We are pleased that he was able to return.
In March 2019 we will be joined by Chloe Robbins. She will be the RA for the HAROLD project, having worked on similar transport safety research for her PhD studies at Nottingham University School of Psychology.
Qi Yao, a visiting scholar, left at the end of 2018 to return to Shenzhen University. Qi worked on a range of projects whilst here, including uniformity metrics, mesopic optimisation and the effect of glare on detection. He worked very hard whilst here and was able to publish several papers.
Zeynep Keskin passed her PhD examination subject to some corrections. Zeynep started by investigating the influence of daylight on seat choice – given the option, do people tend to choose a seat in areas which have more daylight? However, by the time of submission the focus was methods for measuring the influence of seat choice, comparing stated preferences using a questionnaire with revealed preferences (snap-shot occupancy observations and following the routes of individual people).
PhD events: Tentative plans have been made for LumeNet in the coming years. In 2020 it will be hosted by Yvonne de Kort and Antal Haans at Eindhoven Technical University, possibly alongside the next Experiencing Light symposium. In 2022 it will be hosted once again at Sheffield.
Two new interviews have been posted to the BrightLights website – discussions between Kit Cuttle and Antonello Durante (a PhD student at Dublin Institute of Technology) and between Peter Tregenza and Michael Kent.
Discomfort from glare is a perpetual discussion at conferences and in journal papers. The discussion is needed because we are as yet unable to robustly predict the extent of discomfort. One problem is the massive variance found in results – within and between studies.
Together with Michael Kent (University of California, Berkeley – Berkeley Education Alliance for Research in Singapore) and Sergio Altomonte (Université Catholique de Louvain, Belgium) we have been investigating the influence of experimental design. We have demonstrated significant influences of visual task, anchoring, and order.
A further paper, just accepted by Building and Environment, shows the effect of stimulus range bias is sufficient to change discomfort thresholds: so, one way to counter a finding that discomfort is intolerable is to change the stimulus range and it can be reduced to uncomfortable. That is clearly not meant to be a sensible response, but it does describe the current state of discomfort research. A paper describing this research received the best paper award at PLEA2018.
For the 2012 Handbook of Energy I was asked to write a top ten of the milestones in lighting. There were many events and ideas to choose from. This is what I proposed:
The ability for light to be created at will brought about an important change in human life, ending reliance on daylight with its limitations of time and location. The earliest known examples of lamps are about 40,000 years old and these consisted of a wick of lichen or moss soaked in animal fat, placed in the natural hollow of a stone or a shell. These lamps cast dim light but were cleaner and easier to guard than torchwood.
Oil lamp technology
The first significant improvements in the oil lamp came from Argand in the late 18th Century. His tubular wick and glass chimney improved airflow, resulting in a higher burning temperature, a cleaner fire which produced very little soot and smoke. The flame, rather than being the customary orange, was considered to be white and almost dazzling.
Prior to kerosene, lamp oil and candles were made from animal-based material such as whale oil. A process for extracting kerosene from oil was developed in the mid-1800s and its production became commercially viable in 1859 when the first successful oil well was struck in Pennsylvania. Within a year, kerosene had replaced whale oil as the popular fuel and may have contributed to the decline of the whaling industry. Being more economical, kerosene allowed artificial light to be available to wider population.
In the time of elementary navigation aids and imprecise charts, sailing at night-time was perilous; lighthouses provided distant warning of reefs, rocks and sand bars and have saved countless lives. The first offshore lighthouse was constructed at Eddystone, an infamous reef of submerged rocks nine miles off the south coast of England, too distant for shore-based lamps to be effective. Safe navigation required construction of a lighthouse on these very rocks in rough seas.
Smeaton’s Tower, the third of the four lighthouses at Eddystone (1759 to 1877), now stands on Plymouth Ho and is a monument to the heroism of those involved in their construction and of the lighthouse keepers.
We like to use daylight for interior illumination; it is considered to be a better quality light than is electric light, it is usually of a much higher intensity, and it is the zero energy option. The first windows were simply apertures in the external wall. Placing a sheet of translucent material over these apertures meant that daylight could be captured whilst excluding wind, rain and noise. Glazing became common in the windows of ordinary homes in the UK in the early 17th century. Architects are still learning how best to use it.
The first demonstration of electric light was in 1802 when Davy, creator of the miners’ safety lamp, succeeded in making a platinum filament glow, albeit momentarily. In 1809 he demonstrated the first lasting electric light, creating an arc between two carbon electrodes. 1876 saw the first practical carbon arc lamp, developed by Yablochkov and used for public street lighting in Paris. Arc lighting is too intense to illuminate domestic interiors and its use was thus limited to outdoor spaces and displays.
Research of incandescent lighting continued with experiments using filaments made of materials such as carbon, asbestos, platinum and carbonised bamboo. Developments were made by Göbel (1854), Woodward (1875) and Swan (1878) but widespread use arose following Edison who addressed technical issues of filament shape and durability, insulating material, evacuation of the glass bulb, controlled distribution of electricity, and social issues of cost-effectiveness and simplicity for public use.
Visual assessment of ‘how much light’ is subjective – we would each tend to make a different judgement. The need for a visually meaningful and repeatable measure was met in 1924 when the Commission Internationale De L’Éclairage (CIE) adopted the Standard Photopic Observer, Vλ. This is a spectral weighting function that describes how the visual efficiency of the human eye, under certain conditions, varies with wavelength in the visible region of the electromagnetic spectrum.
Regulation of circadian rhythm
Until the late 1900s we were aware of four types of photoreceptor in the human retina: the rods, sensitive at very low levels of light, and the red, green and blue cones, active at higher levels of light typical of daylight and interior lighting. In the early 2000s, a new retinal photoreceptor was discovered, the intrinsically photosensitive retinal ganglion cell (ipRGC), and this appears to connect to non-imaging functions such as regulation of circadian rhythm – the body clock.
Vision at night-time under the dim light of road lighting is known as mesopic. This lies somewhere between the photopic levels of daytime where the cones dominate vision and the very low scotopic levels of a dark night where the rods dominate vision. Mesopic vision is thus a mix of rod and cone vision but until establishment of the CIE Recommended System for Mesopic Photometry in 2010 we were not able to define how this mix varied with light level and lamp spectrum. The coming years will see changes in the way we light roads.
Human factors in lighting
Lighting affects our visual and non-visual capabilities and thus can influence comfort, health, perception and behaviour. Human factors in lighting (1981, 2003, 2014) by Peter Boyce, emeritus Professor of the Lighting Research Centre in New York, is the essential text for the interaction between people and lighting. Boyce managed to capture and distil a vast body of research and he presented this in a way that is informative and enjoyable to read.
Many considerations pertinent to lighting are subjective evaluations – what do people think about their lighting? In his 1977 article 'Quantitative subjective assessments are almost always biased, sometimes completely misleading' and subsequent book Bias in quantifying judgements, Poulton demonstrated that the way in which we ask such evaluations to be made can significantly distort the response, and thus of course the conclusions that are drawn from responses.
The off switch
Light is good. Artificial light has changed the ways in which we live and work. The ability to switch it off at will, to preserve energy or to appreciate darkness, should be viewed with equal relish.
Boyce PR. Human factors in lighting. Taylor & Francis: London. 2003.
Brox J. Brilliant: The evolution of artificial light. Souvenir Press: London. 2011.
CIE 191:2010. Recommended system for visual performance based mesopic photometry.
Illuminating Engineering Society. Discover Lighting! 2011. http://www.ies.org/lighting/history/timeline-of-lighting.cfm
Originally published in: Handbook of Energy volume II: Chronologies, top ten lists and word clouds. Eds. Cutland CJ, Morris CG. Elsevier Science. 2014. pp. 667-668. ISBN: 978-0-12-417013-1.
We have been awarded EPSRC funding for HAROLD: Hazards, Road Lighting and Driving (grant ref EP/S004009/1). This is a £1.2 million project, co-supervised at Sheffield by Professor Richard Rowe, and with a parallel project at Leeds University Institute of Transport Studies led by Professor Natasha Merat (grant ref EP/S003576/1).
The aim of this work is to improve drivers’ ability to detect hazards, with a focus on detecting pedestrians, using road lighting and active visibility aids. The novelty of this work is to consider the effect of distraction on detection; while laboratory experiments tend to allow your full attention given to the task, there are many sources of distraction available to drivers which will reduce their ability to detect hazards.
Work at Sheffield will continue using the scale model apparatus developed for Highways England and these findings will be tested at Leeds using their state-of-the-art immersive driving simulator.
We attended two excellent meetings in August, both promoting good methodology in lighting research. First was the CIE Expert tutorial and workshop on research methods for human factors in lighting, followed by LumeNet. The first event was a series of presentations about specific aspects of research methodology; the second event focussed on ongoing projects by PhD students. An aim of both events was to promote discussion, the ‘confer’ part of ‘conference’ which is often missing from typical conferences due to busy schedules.
Currently there are three PhD students studying daylight-related topics. Zeynep Keskin is just about to submit her thesis on the impact of daylight on seat choice – given the option, does daylight influence where we sit? She has focussed on comparing methods for measuring seat choice.
Choon Yew Chang is investigating window view preference – can we predict the quality of the view from inside to outside? In contrast, Intisar Hussain is investigating privacy, the view from outside to inside, and is investigating window design that permits privacy to the occupant whilst providing optimal daylight exposure.
Five PhD students are studying road lighting, all with a focus on lighting for pedestrians and cyclists. Aleks Liachenko Monteiro is working on reassurance, how we can use road lighting to enhance the feeling of safety, and in her first experiment she evaluated this using Peter Boyce’s day-dark approach.
Yichong Mao and Khalid Hamoodh are considering objective evaluations. Yichong is extending past work on facial emotion recognition by using 3D targets and by considering the performance of this task when conducted in parallel with an obstacle detection task. He recently conducted a study comparing detection of raised and lowered obstacles. Khalid is investing what we look at when deciding whether it is safe to approach another pedestrian, testing the hierarchy of cue suggested by Nancy Clanton.
Hussain Qasem has been studying lighting and cycling and is approaching the submission of his thesis. He has demonstrated that ambient light affects the tendency to cycle; he has shown using eye tracking that there is a tendency to scan the road ahead for hazards, and also that gaze behaviour changes in safer situations. He has also investigated the interaction between road lighting and cycle lighting when detecting approaching hazards.
Scott Fox, who was research assistant for the Highways England project, has spent the past year studying for a masters in research methods, and is re-joining the Lighting Research Group imminently to investigate the influence of distraction on detection.
Qi Yao of Shenzhen University is working here as a visiting research scholar for a year. He is conducting research about mesopic vision, suggesting an alternative model for mesopic vision, spectrum optimisation, and the relationship between CCT and S/P ratio.
Aleks’ study of pedestrian reassurance has led to some interesting thoughts. This study was carried out using the day-dark approach: evaluations of reassurance (perceived safety) were captured in daytime as well as after dark. The benefit of lighting is characterised by the difference between these evaluations: the smaller the day-dark difference, the better the lighting. Mean horizontal illuminance is the widely used metric for outdoor lighting. We found better correlation between reassurance and either minimum illuminance or uniformity than between reassurance and mean illuminance.
Dr Qi Yao (a visiting academic from Shenzhen University, China) is also analysing these field study data, and is working on an alternative uniformity metric that appears to significantly improve the correlation with reassurance and also with direct evaluations of lighting quality. Clearly, this could be a chance finding, so will test our proposals using the field study data gathered by Benedetta Mattoni and Fabio Bisegna of Rome Sapienza University.
Yichong’s first experiment concerns obstacle detection, and he has extended past studies by considering pot holes as well as raised hazards, and by varying the spatial distribution of light. He has finished the first block of trials and the results look promising. Yichong will next add facial emotion recognition to this experiment to consider task performance when two critical tasks are conducted in parallel. This extends past studies by using 3D models of faces of varying expression.
We are joined by a new PhD student, Khalid Hamood, and his first project concerns the visual cues used to evaluate other pedestrians – what is it about them that makes us decide whether or not they are safe to approach? In essence this is testing Nancy’s Rules, an informal hierarchy of anxiety cues suggested by Nancy Clanton, a lighting designer and leader of the IESNA outdoor lighting committee.
Good news for the LumeNet research methods workshop for PhD students – we managed to find industrial sponsorship to cover the conference expenses. So far 25 students have registered, and the aim is to limit this to 40 so that we have good time to focus on each student’s work.
Kiran Maini Gerhardsson, a PhD student in environmental psychology at Lund University, posted two new interviews on BrightLights. One of these is with Mark Rea, who until recently was director of the Lighting Research Center. Mark provides some interesting thoughts on lighting and research.
The majority of our work focuses on lighting for vulnerable road users – pedestrians and cyclists. Jim Uttley recently published two papers showing that, for a given time of day, more people tend to walk or cycle when it is light than when dark, and also that there is a greater risk of being involved in a road traffic accident whilst using a pedestrian crossing when it is dark than when light. This shows that light matters if we want to encourage active travel and make it safe to do so.
We are investigating how safe people feel when walking after dark. Aleks Liachecnko Monteiro carried out a field study of reassurance in residential roads in winter 2017. This was done using the day-dark approach, where the effectiveness of lighting is determined by its effect on the difference between daytime and after-dark evaluations of reassurance, originally proposed by Boyce et al. in 2000. In addition to repeating their method, Aleks is examining alternative approaches to analysis of the rating data, and alternative methods of evaluation (eg eye tracking and involuntary physiological responses).
Yichong Mao is investigating how lighting affects the performance of the more-objective tasks – obstacle detection and facial emotion recognition. Facial emotion recognition requires discrimination between standard facial expressions (angry, happy, neutral) and previous studies have used photographs of actors from independent databases. To allow investigation of spatial distribution of light we will next use 3D faces, and Yichong’s pilot studies so far have been testing models we created using a 3D printer.
Hussain Qasem is investigating cyclists' visual needs. He previously examined how the location and luminance of cycle-mounted headlights, and their integration with road lighting, affected ability to detect road surface hazards. One conclusion was that cycle mounted lamps used for seeing (rather than to-be-seen) should be mounted on the wheel hub rather than the handlebar.
Hussain is now analysing eye tracking data of cyclists in natural setting, specifically comparing results gained using a dual task to identify critical visual fixations with those indicated by involuntary physiological response. Further work on these topics is being carried out with Maria Johansson from Lund University and with Benedetta Mattoni and Fabio Bisegna of Rome Sapienza University.
We carried out an investigation of detection whilst driving for Highways England using a scale model apparatus to simulate the view ahead of a three-lane main road to determine the impact of driving in fog and in the transition between lit and unlit sections of road. Those experiments were conducted by Scott Fox and we are pleased that he has returned to study for his PhD. This will consider the effect of distractions on drivers’ ability to detect hazards. In parallel work with Jan Winter and Stephan Völker from Berlin Technical University we have investigated the visual field of drivers and the influence of glare on detection.
17–22 June 2018. The 16th Light Source conference will be held in Sheffield.
13–17 August 2018. A week of research methods discussion in Copenhagen. The week starts with the first ever CIE Expert tutorial and workshop on research methods for human factors in lighting and the latter two days will be the LumeNet meeting for PhD students.
The second interview has just been posted to BrightLights. Jim Uttley, a research assistant working on road lighting for pedestrians, interviewed Wout van Bommel, whose research in the late 1970s is behind much of current standards for pedestrian lighting.
We have been joined by a new PhD student, Yichong Mao, who will be continuing our recent studies of facial emotion recognition and obstacle detection, but now with a combined task to better represent a pedestrian situation.
Later this month we will be joined for a three-month visit by Benedetta Mattoni, a PhD student from the Sapienza University of Rome, who is working with her supervisor Fabio Bisegna on lighting and pedestrian reassurance after dark.
We have been conducting research with Arup for Highways England about how fog and transition from lit to unlit sections of road affect driving performance. To do this we designed and constructed a 1:10 scale model of a three-lane road in which the test participant sits as a driver with a view of the road ahead.
Scale model apparatus used to investigate detection of obstacles and lane changes.
The ‘driver’ was required to detect two events (sudden appearance of a road surface hazard and lane change of one of two cars ahead) with performance analysed by detection rate and reaction time. A tracking task was conducted in parallel to add cognitive load and to ensure these events were detected in peripheral vision. An LED array permitted changes in the luminance and S/P ratio of road lighting. Artificial fog was used in thin and thick densities as defined by the absorption coefficient.
Two conclusions were drawn from the results:
For driving in thick fog, these data suggest an advantage in road lighting of luminance 1.0 cd/m2 compared with 0.1 cd/m2. Also for thick fog (but not for thin fog or no fog) the lighting of higher S/P ratio produced better (shorter) reaction time for the obstacle detection task (but not for the lane change task, and not for the detection rates of either task). This suggests a benefit of using the higher S/P ratio (1.4 vs 0.65).
These data suggest that when a driver moves from a lit to an unlit section of road, their detection performances decreases near-immediately to that expected for the conditions of the unlit section and that there is no significant subsequent increase in the following 20 minute period in which adaptation may be expected. Similarly, when driving from an unlit to a lit section, there is a near-immediate increase in detection performance that does not change with time.
The experiment was run by Scott Fox who now hopes to carry on this work as a PhD student.
We were recently joined by Aleksandra Liachenko Monteiro, a PhD student. Aleks will be investigating reassurance, specifically, the day-dark approach for analysing whether an increase in illuminance brings a meaningful increase in reassurance. In parallel she is using eye tracking and involuntary physiological responses to support subjective evaluations. In March we will be joined temporarily by Benedetta Mattoni, a researcher from Rome with Fabio Bisegna, who is conducting similar studies.
Later in the year we will be conducting tests to see how past results of obstacle detection and facial emotion recognition are influenced by variations in glare and light source location, the two objective tasks suggested by eye tracking to be most critical for pedestrians. Hussain Qasem is currently analysing further eye tracking data recorded with cyclists to investigate their critical tasks.
The fourth Velux Daylight Academic Forum will take place in Berlin on 4 May, organised by Jens Christoffersen. This is a methodology workshop for PhD students of lighting, similar to LumeNet (which has a broader focus in lighting, next scheduled for summer 2018). Students have the opportunity to present for criticism their proposed research methods, with feedback given by senior researchers. This can be a tough process, but far better to gain critical feedback before an experiment starts than as review of a submitted paper. If you are studying daylight, it would be crazy not to go.
As a new researcher, it can be difficult to meet the senior researchers, the people whose papers you have read but always seem to be busy at conferences. I set up BrightLights to help. This is an opportunity to interview a person you want to meet; record and transcribe your conversation and then send it to me with a photo. The first post is a conversation between Kynthia Chamilothori (EPFL, Lausanne) and Peter Boyce – it’s an interesting read.
LumeNet 2016 was organised by Peter Hanselaer and held at KU Leuven Technology Campus in Ghent. Over the two days (7–8 April) 30 PhD students presented their work, seeking feedback from the experts. Special thanks must go to Kevin Houser who travelled all the way from Penn State to attend.
For 2017, Jens Christoffersen will organise the Academic Forum, which focuses on daylight research. LumeNet 2018 is set for August 2018 in conjunction with a CIE expert symposium on research methods, and will be held in Copenhagen.
Students and reviewers at LumeNet 2016
KU Leuven Technology Campus
Hussain Qasem has just finished trials recording the eye movements of cyclists: this was done in parallel with a dual task and skin conductance recordings from which we hope to identify critical visual fixations.
Zeynep Keskin enters the final year of her PhD studying daylight and seating preferences: her main challenge is to link perceived influences with actual behaviour.
Holly Castleton left after a two-year research post to take a lectureship at Sheffield Hallam University. In February we were joined by Scott Fox for a Highways England research project investigating how fog affects driver’s ability to detect hazards.
Six of the Lighting Group attended the 28th session of the CIE in Manchester, presenting nine papers between us on pedestrians, cyclists, research methods, and non-visual effects in schools and classrooms. We attended the first meeting of the CIE technical committee collating empirical evidence of lighting for pedestrians and proposed a working group be set up to focus on cyclists.
Steve Fotios was awarded the 2015 CIE Waldram Gold Pin for Outstanding Contribution in Applied Illuminating Engineering – a rare distinction. Steve was a member of the organising team for this conference; one achievement was to source ethical conference bags – manufactured from organic cotton by an organisation that provides work and support for women previously trapped in abusive circumstances.
Steve Fotios attended the first meeting of the IESNA outdoor lighting committee in Kalamazoo. The aim is to update guidance for public spaces, and with a committee combining researchers, designers and equipment producers, it is well placed to do so. Steve is also working with CIE and ILP committees on pedestrian lighting so this is an opportunity to develop widely-informed and consistent advice.
The IESNA Outdoor Lighting committee (Left to right): Randy Burkett, Naomi Miller, Steve Fotios, Charles Stone, Nancy Clanton, Susanne Seitinger, Chip Israel and Rick Utting.
In the past months both Jemima Unwin and Biao Yang submitted their PhDs and both were successful with only minor corrections. Jemima and Biao were part of the MERLIN team studying lighting for pedestrians, Jemima investigating the reassurance gained from lighting and Biao how lighting effects our ability to judge the intention of others. Both made a great contribution to lighting research.
Biao is now working as a researcher on an eye-tracking project at UCL and Jemima is a lighting design tutor at Brunel in addition to her own design practice.
Two new students have joined the School of Architecture to study lighting and human factors. Hussain Qasam is investigating lighting for cyclists, extending our recent focus on pedestrians, and in the first place is doing an obstacle detection test using the simulator developed by Jim Uttley and Chris Cheal.
Alberto Urrutia is investigating lighting in prison environments, in particular whether enhanced ipRGC stimulation is an effective in reducing low level disruptive behaviour – this will be a good parallel to Andy Colau’s study of non-visual effects of lighting in schools.
April 2014 was a busy month for lighting research conferences.
IESNA Light and Behavior Conference (Cleveland, USA)
At this event, designers are invited to discuss their lighting strategies and researchers are then invited to respond to this: what strategies are supported by research, what ideas are not supported by research, and what further research is needed? Along with Jack Naser I was invited to provide research response on design for urban environments presented by Nancy Clanton and Randy Burkett.
LumeNet (Berlin Technical University)
LumeNet is the workshop for PhD students of lighting that I established in 2011 with Jens Christoffersen. The objective is that PhD students receive critical feedback on their proposed aims and methodologies at an early stage in their work, while there is still time to influence the work, providing alternative guidance to that which they might receive from their own supervisor before carrying out an experiment.
The senior researchers giving this feedback are carefully chosen. They are leading names in lighting research, and these are people with whom the students might not otherwise have the chance to engage. LumeNet was held at Sheffield in 2012 and the next event (2016) will be in Sheffield, Copenhagen or Belgium – still to be decided!
CIE 2014 Lighting Quality and Energy Efficiency Conference (Kuala Lumpur)
This is a major annual event in the field of lighting. Five researchers from Sheffield attended, presenting four papers and attending the following technical committee meetings which work towards guidance reports and standards.
IMPROVE (Lund University, Sweden)
This was a one-off event, bringing together a group of invited speakers associated with lighting and vision but without any intentional theme other than good research. The aim was to share skills between people coming from different backgrounds – environmental psychology, behaviour, medical and more. An excellent idea, there should be more events like this.
Students and reviewers at LumeNet 2014
Berlin Technical University
Two new PhD students have joined the research group. Andy Colau is investigating how lighting may be used to improve students' learning, enhancing stimulation of the circadian system to improve alertness particularly in the morning. Zeynep Keskin is studying daylight. Along with current students in the group they will be attending LumeNet 2014 in April 2014 at Berlin Technical University, organised by Stephan Völker and Martine Knoop.
We have also been joined by Holly Castleton, a research assistant. Holly’s primary role is to work on research funding. Being a building engineer she is new to lighting research and to gain direct experience is running a facial expression recognition test, extending the work of Biao Yang using a new apparatus to enable coloured images to be used at low light levels.
Deniz Atli has finished analysis of her spatial brightness experiment. This was the first study to repeat the full-field sequential discrimination experiment of Berman et al. (1990) and for further validity we included a second procedure (matching), an additional null condition, added surface colour as a variable, and used a third lamp to compare the effect of gamut area in addition to S/P ratio.
Jemima, Biao and Jim have made excellent progress in research of lighting for pedestrians.
Jemima’s analysis of on-street safety ratings using the day-dark approach of Boyce et al. (2000) may lead to new proposals for an optimum illuminance.
Jim’s extensive analysis of the eye-tracking records was able to demonstrate the benefits of the dual task approach and he is now setting up a pedestrian simulator to extend our research of obstacle detection.
Biao completed his second experiment of facial expression recognition, gaining 34,560 judgements from the 20 test participants for 24 target faces, with 18 combinations of luminance and lamp, and 4 combinations of distance and duration.
Deniz Atli and Naoya Hara are setting up the apparatus to repeat the Berman et al. 1990 study of lamp spectrum and spatial brightness. Berman et al. used two lamps of near-identical chromaticity but different S/P ratios, seeking to identify spectrum-based effect on spatial brightness other than the chromatic contribution.
The resources available mean we are using a scale model (booth) rather than a small room for the evaluations, as did Berman et al., but this still provides full-field stimulation. While our LED array does not permit replication of the precise stimuli used by Berman et al., we have retained the approach of near-identical chromaticity but different S/P ratios.
The procedure also includes a true null condition (two sources of identical SPD and luminance) and will be repeated with achromatic and coloured environments. In parallel, the same stimuli will be used in a sequential matching task, thus permitting evaluation of methodology, with a third source chosen to evaluate effects of gamut area at constant S/P ratio.
Lighting for pedestrians
Jim Uttley used a dual-task approach to aid identification of critical visual tasks from the results of eye-tracking, this carried out at daytime and after dark. The results suggested that the near path and distant pedestrians were the most frequent fixations at instances when a slow response to the dual task suggested attention was focussed on something critical. The video records are now being further interrogated to explore questions of eye-tracking methodology and issues related to inter-personal judgements.
Biao Yang has completed a study of lighting and interpersonal judgements. This examined how luminance (0.01, 0.1 and 1.0 cd/m²), interpersonal distance (3 fixed distances , depending on task) and lamp type (HPS and MH) affected ability to judge emotion from facial expression or body posture, and whether or not eye-gaze was direct, and this was done using a 1000 ms exposure.
The results suggest a plateau-escarpment relationship between luminance and performance and this gives one approach to identification of suitable light levels: a luminance of 0.1 cd/m² is required for accurate recognition of facial expression at 4m and body posture at 10m and 30m, while a luminance of 1.0 cd/m² is required for accurate recognition of facial expression at 10m and above-chance level of recognising gaze direction at 2m. Biao will next carry out a second experiment to validate these conclusions.
Jemima Unwin is exploring reassurance – the subjective evaluation of safety. In her first study she asked whether the presence of road light enhances reassurance using an open-ended interview method. This confirmed that people are happier to walk alone after dark when road lighting is present, and that this tends to be a more-obvious factor than are spatial features that might be associated with prospect or refuge.
Jemima is now working on a study evaluating the influence of light level – does higher illuminance enhance reassurance – using the procedure reported in Boyce et al., 2000.
This has been a busy year for the Lighting Research Group, with visits to present work at the Lux Pacifica and CIE Mid-Term conferences and planning ahead for Lux Europa. A workshop was held at the CIE meeting to discuss evidence behind road lighting design criteria, repeating the event held in Hangzhou, 2012. The outcome is a proposed new TC in Division 4 to report on empirical evidence for lighting and pedestrian tasks.
Jim Uttley has completed the eye-tracking study and now has a large set of data to work with. His first task has been to identify the likely critical visual fixations, as noted by delayed responses to the dual task. He will then use the standard approach to analysis (all fixations) to investigate differences.
Biao Yang is also analysing the video records to identify the probability of fixating on other pedestrians in the field of view and the distances at which fixations occur, as this will contribute to his study of inter-personal judgements.
Jemima Unwin has completed one set of reassurance (perceived safety) studies in residential streets and is planning two further trials (summer 2013, winter 2013) to better understand the influences on reassurance judgements.
The MERLIN project focuses on lighting for pedestrians. I am now working towards extending this to include lighting for cyclists, and would be pleased to hear from collaborators with ideas!
Work on lamp spectrum and spatial brightness is continuing: we are setting up a test room to repeat the experiment reported by Berman et al. (Lighting Research and Technology, 1990) and this will be done in separate trials by Naoya Hara and Deniz Atli enabling us to confirm Berman’s results, investigate cultural differences, and test different proposals for brightness effects of lamp spectrum. Thanks to Professor John Barbur and Gary Bargarry at City University for the design and supply of a tuneable LED array.
The LumeNet PhD conference was the result of my long-term ambition being met with enthusiasm from Jens Christoffersen at VELUX. Two events have taken place to date, with the daylighting academic forum scheduled for mid May 2013 and the next LumeNet in Berlin in April 2014. These are essential events for PhD students of lighting because of the opportunity to get critical feedback from leading international researchers. I have also been pleased to see other conference events now including a PhD focussed session, including Experiencing Light 2012 and the CIE Mid-Term conference 2013.
The Lighting Research Group has been joined by two new members. Dr Naoya Hara is a visiting researcher from Kansai University, Japan and James Uttley is a PhD student with a background in psychology. Initially they are working together to investigate the visual tasks of pedestrians after dark – what are the important objects they look at?
This is being done using eye-tracking (thanks to MERLIN partner Peter Raynham at UCL for the loan of this apparatus) with a dual task being employed to help reveal the important objects of attention from within general observation.
Good progress has been made within the MERLIN project. Jemima Unwin has completed her investigation of 'Does lighting matter?' with regards to pedestrian reassurance and is now setting up a second study to identify the relationship with light level – a repeat of the procedure reported in Boyce et al. (Lighting Res. Technol., 2000. 32, 79-91) but using residential streets rather than car parks.
Biao Yang has completed two pilot studies regarding the inter-personal judgements made between pedestrians, one exploring the features we might extract about other people at different distances, and one exploring whether facial expressions and body postures lead to consistent judgements of threat/non-threat. He is now planning a new series of tests examining how judgements of gaze direction, expression, intent, and facial recognition are affected by lighting at low levels. We hope to be presenting results of these studies at the Lux Pacifica, Lux Europa and CIE conferences in 2013.
A recent project explored metrics for predicting the effect of SPD on brightness at mesopic levels. This work resulted in new national guidance for subsidiary streets as reported in the Institution of Lighting Professionals new report PLG03 Lighting for subsidiary roads: Using white light sources to balance energy efficiency and visual amenity.
The aim of the MERLIN project is to investigate how design light levels in residential roads might be set if based on visual tasks. Steve Fotios led a workshop at the recent CIE conference in Hangzhou and found
general agreement that the basis behind current design standards is very tenuous, and
support for setting up a new technical committee to report on approaches to setting design standard for road lighting.
LumeNet 2012: Thanks to everyone who attended, it was a great success!
There were 41 students (from Austria, Belgium, Czech Republic, Denmark, Finland, Germany, Ireland, Italy, Netherlands, Norway, Poland, Spain, Sweden, USA, and the UK) who presented the methodology of their work for review by invited senior researchers - Peter Boyce, Jens Christoffersen, Steve Fotios, Kevin Houser, John Mardaljevic, Mike Pointer, Jennifer Veitch and Stephan Voelker.
Thanks to VELUX, Thorn Lighting, Zumtobel Lighting and the Society of Light and Lighting for sponsoring LumeNet 2012. With their support it was possible to avoid a registration fee, which appears to be a barrier to attendance at some events.
Comments received from the students and reviewers show that it was considered to be a worthwhile and constructive event.
I thought it struck just the right balance between detailed discussion of projects and general discussion of process. I am sure everybody learnt something and many made valuable contacts. Overall, it was a very worthwhile meeting.
- Peter Boyce
An excellent meeting which seemed to have been enjoyed by all. The organisation was brilliant with a good balance between work and play.
- Mike Pointer
I loved the conversational format, which allowed our small groups to pursue ideas without the need to stick to a rigid timeline. We were able to linger on a topic when deeper discussion was warranted, and these were some of the most fruitful periods of the workshop. It was also great to learn about the research problems that others are studying, and it was exciting to see so much enthusiasm and passion from all involved.
- Kevin Houser, Penn State University
Thanks very much for organising it, it was a really worthwhile event - in fact I'd go so far as to say it was a 'must do' for any PhD student in the field of lighting. I'm particularly grateful for being able to have some time with Peter Boyce!
- Ruth Kelly, De Montfort University
I certainly found the entire event beneficial at this very early stage of my research. It opened my eyes to the standards that I need to achieve and gave me new found enthusiasm to work towards them. Thank you once again for your hospitality and critical feedback on my work.
- James Duff, Dublin Institute of Technology, Ireland
I think I met all the goals and I'll recommend Liisa to send someone there also in the future. I was really impressed for your devotion on the lighting.
- Heli Nikunen, Aalto University
We have recently made good progress on four projects. Three of these relate to the MERLIN project investigating lighting for pedestrians.
Work has progressed in characterising the effect of lamp spectrum on spatial brightness at photopic levels. We have finalised the review of past studies of spatial brightness to identify the evidence that gives reliable estimates of the magnitude of the spectrum effect.
The first stage of analysis has been to use data from matching and discrimination tests (as these give the magnitude and direction of the effect) to screen potential metrics for brightness, eg the S/P ratio (the basis of Berman’s brightness lumens), gamut area, colour metrics such as CCT and CRI, and models of equivalent luminance.
Initial analysis suggests ratios of gamut area and ratios of (S/P) 0.5 give the better predictions of the metrics examined: perhaps a combination of these two will show an improvement. Any metrics proposed will next be validated by experiment, using side-by-side matching and separate-evaluation rating procedures.
LumeNet 2012 is rapidly approaching and we are making plans to ensure it is an interesting and worthwhile event for the 40 PhD students we are expecting. Looking forward to meeting all who attend.
The School of Architecture has recently moved backed to the Arts Tower and a first task for the lighting group is setting up the laboratory rooms. Biao Yang has started work on his PhD regarding judgement of intent: what is it that a pedestrian uses to interpret the intent of other pedestrians? We suspect it is more than simple facial recognition. He has already found some interesting ideas, for example, questioning the basis of Hall's personal proximity limits.
Jemima Unwin has carried out a pilot study of reassurance , our expression for perceived safety or fear of crime, and from this is revising the procedures for a principal study that will take place in the coming winter. Her main question is: does lighting contribute to reassurance? Some past studies have suggested so, but in the context of other social and environmental factors the effect is questionable.
Chris Cheal is setting up the obstacle detection apparatus to examine the effect of luminance on detection probability: the pilot study results suggested a plateau-escarpment relationship, as exists with RVP at higher light levels, but with only three light levels in the pilot study we must question whether it was an experimental artefact. If the plateau-escarpment transition is real, however, this would provide a useful clue as to minimum light levels for pedestrian lighting.
Deniz Atli has had a successful first year in her PhD studying spatial brightness. She has completed a review of studies using the category rating method, along the way publishing articles on response-points in rating scales and a comparison of brightness and clarity ratings, and is now starting a regression analysis of metrics which might predict how lamp spectrum effects spatial brightness.
LumeNet 2012: This is a conference/workshop for PhD students studying lighting – themes within lighting, daylighting, colour and vision are welcomed. Students will present their work, in particular the methods they intend to use for experimental work, and will receive critical feedback from a group of experts who will attend the event. Experts include John Mardaljevic, Jens Christoffersen, Peter Boyce, Mike Pointer and Jennifer Veitch.
LumeNet takes place on 19–21 June 2012 at the University of Sheffield, please register soon if interested. Thanks to VELUX, Thorn Lighting, Zumtobel Lighting and the Society of Light and Lighting for their sponsorship.
March is the official start of the MERLIN project at Sheffield and we have been joined by a new PhD student, Jemima Unwin, who will investigate reassurance in streets at night-time – does lighting have any effect and, if so, what characteristics matter?
Jemima is an architect who has international experience having worked for practices in Kyoto, Sydney, Manchester and London. Her architectural education was completed at the Mackintosh School of Architecture in Glasgow and Ecole d'Architecture de Paris-Belleville in Paris.
Her long-term interest in both electric and day lighting was formalised in the completion of the MSc in Light and Lighting at UCL in 2010, after which she was awarded the prestigious Millar Award by the Worshipful Company of Lightmongers. She is excited to join such a fascinating project which involves understanding how lighting affects people’s experience of their urban environment.
A second PhD student, Biao Yang, is expected to start in summer 2011. Biao is currently completing his masters study with Yandan Lin at Fudan University. His work will investigate recognition of facial expression under different types of street lighting. Previous work in this area has tended to investigate the identification of faces and has allowed continuous scrutiny of the target face. This is probably not the right approach.
The task for pedestrians is more likely to be determination of whether the person is friendly or not: the psychology literature suggests investigation of expression rather than identification may give a better clue as to whether or not the person should be avoided, and early results from the eye tracking survey at UCL (Peter Raynham and Navaz Davoudian) suggest that the scanning of a face is carried out in a very brief time rather than continuous scrutiny.
A new methodology has been proposed (Fotios and Raynham, Lighting for pedestrians: Is facial recognition what matters? Lighting Research & Technology 2011; 43/1: 129-130).
Tharinee Ramasoot submitted her PhD thesis 'Investigation of lighting and disturbing reflections on display screens: A new model for judging acceptability; this was examined by Peter Boyce and was accepted with only very minor corrections. Well done Tharinee! Her work contributed to the revised SLL guidance for lighting in classrooms. She has now returned to Silpakorn University, Bangkok to resume her lectureship.
We have been joined by a new PhD student, Deniz Atli from Turkey. Deniz previously studied at Bilkent University, completing her undergraduate degree in the Department of Interior Architecture and Environmental Design and her masters degree in architectural lighting.
Her masters thesis examined the effects of colour and coloured light on depth perception, comparing different colour combinations of backgrounds with objects, and with this focus on visual perception she is an ideal new colleague for the research group at Sheffield. Deniz will be studying lamp spectrum and brightness perception in interior spaces.
Ongoing projects: Development of a new method for measuring facial recognition under different road lighting conditions, examination of the method of adjustment (ie dimming) for setting preferred levels of illuminance or CCT.
We received the good news from EPSRC that the MERLIN project (Mesopically Enhanced Road Lighting: Improving Night-vision) was approved for funding. This is a collaborative project with John Barbur at City University and Peter Raynham at UCL with a total value of over £1 million. Our intention is to find out what visual tasks are critical for pedestrians at night-time, how these tasks are affected by the amount, spectral distribution and spatial distribution of light, and thus set design targets for pedestrian lighting.
PhD studentships – The funding awarded to the University of Sheffield for MERLIN includes two PhD studentships covering fees (UK/EU residents) and a student bursary for three years. These are likely to start in February and June 2011. Please contact us if interested.
It has been a busy start to the year, with presentations delivered at Stockholm Lighting Days and the CIE 2010 conference in Vienna, a new model developed to predict disturbing reflections on display screens, new data emerging on spatial brightness at mesopic levels, progress with the CIE and ILE committees tasked to consider spectral power characteristics of lighting at mesopic levels, and a new technical committee set up within the CIE to examine research methods.
In February, Navaz Davoudian's PhD thesis was examined by Professor Stephan Völker of the Berlin Institute of Technology. We are pleased to report that she passed, subject only to minor corrections.
Tharinee Ramasoot has developed a new model for predicting the threshold luminance to avoid disturbing reflections on display screens. This model is based on results of an experiment using luminance adjustment to find the disturbance borderline; the experiment used seven different screens including interactive white boards, three sizes of reflection, and two viewing locations.
The results were validated using a second methodology (category rating at a series of fixed luminances) and through comparison with previous data and models. Tharinee is due to submit her PhD thesis in June 2010 and this will be examined by Peter Boyce.
The brightness matching trials being carried out by Chris Cheal are nearly complete. These trials are carried out at mesopic levels, providing data to aid the selection of optimum characteristics for lighting in residential roads. A critical issue these data reveal is that some lamps which create spaces that appear to be very bright do not give pleasing appearance of skin or a colour array; a compromise is needed.
BS5489-1:2003 allows the illuminance of lighting in residential roads to be reduced by one class of the S-series when using lamps of Ra>60. These new brightness data, together with our previous work on obstacle detection, suggest the illuminance reduction should be allowed for lamps having both Ra>60 and S/P>1.6 (this is a first estimate, to be reviewed on completion of trials). Ra correlates with the preferred appearance of skin and colours and the S/P ratio correlates with brightness judgements and visual performance.
In January we were pleased to receive a visit from Ásta Logadóttir of the Danish Building Research Institute. She is carrying out a study of preferred illuminances and CCTs in an office setting and this meeting gave us a chance to share ideas on methods.
A new technical committee has been set up within the CIE, TC 1-80: Research Methods For Psychophysical Studies Of Brightness Judgements. The terms of reference are:
To report on research methods (both research design and statistical analysis) for psychophysical studies of spatial brightness judgements. The aim is to bring best practices from psychology into the wider awareness of people in the lighting community who wish to use such tools in their own work, to avoid errors that plague the existing literature.
Steve Fotios is the chairman of this committee - please contact him if you would like to participate.
Navaz Davoudian is about to submit her PhD thesis; she has examined how the saliency of an object is affected by the background pattern of lighting. If lighting master plans were able to control background lighting this would enable luminance contrasts to be reduced for the same level of saliency. Navaz has recently started a new research post at UCL, working with Peter Raynham.
Tharinee Ramasoot is working through the results of the screen glare tests. She gathered an immense amount of data so this analysis will take a while to complete. The results show that there is a wide variance in the luminaire luminance that can be tolerated by different screens before the reflections become disturbing. A new system of guidance may be needed to deal with this, and to future proof guidance against further changes in DSE technology.
Chris Cheal has set up the brightness matching apparatus. We have chosen five different lamps for these trials, these being used in all possible pairs, and this will enable predictions of brightness based on conventional lamp colour metrics to be tested. The tests will mainly use the brightness matching procedure backed up by a forced choice judgement at equal illuminance; foveal visual performance will be tested using a low contrast acuity chart, and preference of skin appearance will be examined.
Steve Fotios is chairman on the new ILE Mesopic Vision panel. This has been set the task of providing road lighting designers with guidance on application of new research of mesopic vision. He is also setting up a proposal for a new CIE technical committee to examine research methods for subjective evaluation of lighting.
Brightness of street lighting
Over the past year we have carried out brightness matching trials to verify the results we gained from a previous study using side-by-side matching. In one set of trials we repeated the side-by-side match but used a range of different field designs, from a neutral uniform surface to an interior space with coloured surfaces. There was little effect on the illuminance ratio at equal brightness.
A second study used sequential evaluation (lighting from two different lamps used in temporal succession to illuminate a single space) rather than the simultaneous (side-by-side) evaluation used previously. There is negligible difference between the illuminance ratios at equal brightness derived from the two methods.
These findings provide support for continued use of the side-by-side matching task. We are now setting up a series of trials to identify a lamp quality metric for specifying when it is acceptable to reduce the illuminance of street lighting.
Three methods have been used to evaluate the effect of light sources reflected in display screens including interactive whiteboards. Two of these were subjective evaluations, using an adjustment task and a rating task to identify the luminance of the disturbance threshold. The third was a reading speed task, to provide an objective measure of the effects.
Analysis of the results to date shows a wide variation of disturbance threshold luminances between different types of screen, and that these luminances are not well specified in current British Standards.
We are proposing an alternative system, in which minimum qualities of display screen are specified to meet the characteristics of the installed lighting rather than vice versa as is the current situation. Specifying minimum qualities of display screens rather than limiting qualities of the lighting provides a way of future-proofing lighting guidance.
Alison Chang was awarded her PhD thesis for a study of spatial attributes of buildings and reassurance when wayfinding. We intend to extend this work to investigate effects of lighting on reassurance for pedestrians at nighttime.
Chris Cheal has constructed a new apparatus for sequential brightness matching tests. We will be using this to see how the results compare to previous matching tests using simultaneous evaluations (side-by-side presentation). The recent study of field design suggested the addition of coloured surfaces or objects into the visual field does not have significant effect on the matching task.
Tharinee Ramasoot is setting up the screen reflection apparatus to use the rating method of subjective evaluation, and these results will be used to confirm those previously gained using the adjustment task.
Steve Fotios published an article in the IESNA journal Leukos discussing the category rating method for judging the brightness of lighting. Many studies have used this method to compare lighting of different spectral power distribution and this article offers an explanation for why some studies report a significant effect yet others suggest negligible effect.
RAE2008: The School of Architecture performed very well in the 2008 Research Assessment Exercise. Within the 'Architecture and the Built Environment' unit we are ranked joint first or joint third (depending on the method used to interpret the quality profile) out of the 35 universities who submitted.
Tharinee Ramasoot has finished the first stage of the screen reflection tests, an adjustment task using younger subjects and seven types of display screen. These initial results show that some screens can tolerate much higher luminaires luminances than are permitted under the current systems of specification.
Chris Cheal is conducting tests to examine how the brightness matching task is affected by design of the illuminated field, and these range from a uniform surface to a space furnished to simulate a residential street. This preliminary work is needed to guide the design of experiments we will conduct to compare spatial brightness from lamps of different SPD.
Two series of tests are about to start: evaluations of reflections on display screens and brightness matching at mesopic levels using a wide range of lamp SPD. Both of these are EPSRC funded projects.
Presentations were given by group members at several conferences through the summer – Balkan Light 2008, PLEA 2008, ILE annual conference and the SLL/LR&T seminar. Steve Fotios will be attending the IESNA annual conference in November to work with the visual effects of lamp spectral distribution committee. Steve was chairman of the team drafting ILE technical report TR29: White Light, and this was launched at the ILE annual conference.
The obstacle detection research found that lamp type affects the detection of pavement surface irregularities at low illuminance (0.2 lux) but not at higher illuminances (2.0 and 20 lux), and for the three lamps tested the effect is as predicted by S/P ratio – the higher the S/P ratio, the better the ability to detect small obstacles. The final report will be available from this website.
Professor Peter Tregenza, Professor Peter Boyce and Andrew Gooding of Urbis Lighting attended an ideas workshop at the University of Sheffield to discuss the new research projects – Lighting for the classroom of the future and a metric for specifying white light. This meeting lead to interesting ideas for improving the proposed research strategies; for example, the street lighting tests will now include assessment of skin appearance under different lighting in addition to the proposed brightness and visual performance tests.
Steve Fotios has commenced drafting the ILE technical report providing evidence in support of white lighting in subsidiary streets. He is chairman of the panel that also includes Nigel Townsend (Urbis Lighting), Kevin Moss (West Sussex County Council), Mark Hooper (Lighting and Electrical Design Services Ltd) and Alistair Scott (Design for Lighting). The report is due to be published in time for the ILE annual conference in Bristol, September 2008.
EPSRC funding for two research projects was awarded to Dr Fotios:
Lighting for the classroom of the future: Acceptability of glare (EPSRC ref EP/F029276/1)
Street lighting: A metric for specifying white light (EPSRC ref EP/F035624/1)
Chris Cheal was awarded his PhD for a thesis on street lighting and lamp SPD. It was examined by Peter Raynham of UCL.
Steve Fotios received the Institute of Lighting Engineers (ILE) award for best presentation at the 2006 ILE National Conference. This presentation discussed the use of whiter light sources for lighting in residential streets, a project funded by EPSRC.
Steve Fotios has been nominated as Chairman for the brightness sub-committee of the IESNA (Illuminating Engineering Society of North America) Effects of Lamp Spectral Distribution committee. The committee constitution has been revised by the chairman, Brian Liebel, to investigate SPD effects at photopic conditions whereas it was previously concerned only with mesopic conditions.
Project agreed with Colette Knight, Philips Lighting Ltd, to investigate obstacle detection under different types of light source. Obstacle detection is one of the primary visual tasks of pedestrians, and has not previously been directly tested.
Received the Leon Gaster Award 2006 from the Society of Light and Lighting for an article in Lighting Research and Technology. Fotios SA, Cheal C and Boyce PR, Light source spectrum, brightness perception and visual performance in pedestrian environments: a review, Lighting Research and Technology, 2005; 37(4); 271-294.