MERLIN: Mesopically Enhanced Road Lighting: Improving Night-vision

This project will investigate how the lighting of roads in residential areas might be changed so as to preserve the benefits of good vision while minimising energy consumption. In residential roads the road lighting is designed primarily to meet the needs of pedestrians. Benefits of road lighting for the pedestrian are enhancement of safe movement (e.g. increased visibility of potential trip hazards) and perceived safety. Minimising energy consumption will also reduce operating costs and carbon emissions. Road lighting in the UK consumes 2.5TWh electricity per annum, of which approximately one third is associated with residential streets (the remainder being trunk route lighting, signs and bollards). At a typical unit cost of 0.10/kWh this represents an annual cost of over 83 million. Recent advances in lighting technology and in our understanding of mesopic vision have the potential to improve specification of the spectral power distribution (SPD) and spatial distribution of lighting to achieve reductions in the level of illumination, and hence a reduction in energy consumption, whilst maintaining or even improving the level of benefits. A direct route to improved lighting design and thus energy savings will be through better specification of residential street lighting criteria. This will be realised by amendments to British Standard BS5489-1: two of the applicants serve on national and international standardisation bodies for road lighting and so are able to implement such amendments.Comparison with international practise suggests that light levels used in the UK may be excessive. In the UK, residential streets are lit to average pavement illuminances in the range 2 to 15 lux. However, Australia tends to use average illuminances of only 0.5 to 0.85 lux, and 3 to 5 lux in Japan. Thus there is clear potential to reduce light levels in the UK.The project involves three major stages. First, we propose to identify and characterise the principal visual tasks of pedestrians at night-time and the key aspects of the visual environment they consider to be important. This will be established by analysing the pedestrian environment on current residential roads: an eye tracking study will be carried out to determine what objects are viewed when walking at night-time together with a reassurance study to determine how the spatial and photometric characteristics of streets affect perceived safety. To our knowledge, the fundamental question of 'What is important to pedestrians' has not been the focus of any previous research. The work proposed for the second stage will identify how these critical visual tasks are affected by the level of illumination, spectral power distribution and spatial distribution of lighting. This will be examined in two parallel studies; (i) an examination of threshold visual responses, and (ii) an examination of the performance of applied tasks, such as facial recognition. The final stage will determine the optimum criteria for lighting design in residential streets and hence the optimum illuminances and light source properties based on the data obtained in the project. We will subsequently demonstrate and validate the benefits of implementing the new findings to end users by re-lighting a sample of residential streets using the optimised design criteria that will emerge from this project. Because this work will be dependant upon the previous findings and because it requires agreement with local authorities this part of the project has been scheduled as a follow-up activity. Agreement has already been obtained from lighting equipment manufacturers to donate much of the equipment needed for this work.

This project has the potential to transform lighting in residential streets and improve energy efficiency worldwide. The primary characteristics of street lighting are the illuminance (light level), the spatial distribution of this light, and the spectral power distribution (SPD) which defines both the colour appearance of the lighting and the colour rendering of illuminated objects. This proposal will develop new approaches to research on visual responses at low light levels which aim to provide a more useful description of how vision changes in order to determine the optimum illuminance, SPD and spatial distribution of lighting to meet pedestrian needs. Emerging light source technologies such as LED arrays and electrode-free metal halide lamps now offer the ability to tune the SPD of lighting to implement the findings of vision research - this has only been partly possible with conventional lamps. Improved knowledge of fundamental visual capabilities at low light levels will enable lighting design criteria to be optimised, so that the visual benefits of lighting are maintained whilst the energy consumption is reduced. Local authorities will thus be able to reduce the 83 million annual expenditure on street lighting. Optimisation means that the spectral power distribution and spatial distribution of lighting will be chosen such that the average illuminance can be reduced: avoiding excessive illuminance also offers a decrease in light pollution - light reflected to the sky at night. An important aspect of this project is to consider pedestrians visual needs in relation to aging by examining subjects in different age groups. Factors such as parallel processing of visual information that is directly related to age and the effects of scattered light, that are also known to affect visual performance in the older population, will be examined. British Standard BS5489-1:2003 provides guidance on the selection of lighting design criteria. This guidance is based largely on convention - continuing existing practise because it appears to work OK. Experimental research tells us that there is much scope for improvement in this guidance, but what is lacking is the means of generalising this research in a way that can be applied in practice. An objective of the proposed work is the provision of improved evidence for this standard. The optimum spectral characteristics will be determined in tests using a tuneable LED array. Using this apparatus avoids the limitation of using only conventional lamps, these offering a limited range of SPD possibilities. The resultant findings will therefore be independent of the lighting technology involved and will outlast and may even lead to further technological developments. Such data will also provide targets for lamp manufactures to design lights that fulfil the optimum qualities of lighting and luminous efficacy. Current guidance for street lighting is based on the pedestrian visual needs suggested by Philips Lighting in the 1980's and these have governed the design of previous experimental work, e.g. facial recognition, visual orientation and obstacle detection. There has been no check to see whether these needs are valid and of their relative importance. The first stage of the proposed work involves an evaluation of what matters to pedestrians at night time, and we anticipate this will provide a focus for international research. This project will provide new fundamental data of mesopic vision with relevance to the way pedestrians use streets at night. Improved understanding of visual performance in the mesopic range will also contribute to current efforts to develop and standardise scales for mesopic vision.