Road Luminance & Illuminance Dimming Level (% dimmed) L1
The Quest for Safer Roads: LED Lighting, Dimming and Driver Performance Professor Joanne Wood PhD Dr Gillian Isoardi PhD Queensland University of Technology
QUT Vision & Driving and Lighting Research Professor Joanne Wood; Dr Alexander Black •
Vision Science Research Group, QUT
Dr Gillian Isoardi; A/Professor Ian Cowling •
Lighting Research Group, QUT
Pilot study supported by: • Dept. Transport & Main Roads • LED Roadway Lighting
SLSC Roadmap 2016/17: on Research Few published studies worldwide have investigated the relationship between road lighting levels and crash rates at night Examples of Potential Research Projects: LED lighting application and road safety effectiveness Substantiate road safety benefits of LED (white light) Provide evidence for acceptable dimming conditions
‘Robust, credible research should underpin all policy measures’
Road Lighting and Safety Design principles of streetlighting have changed little since its inception around the 1930s, however: • Drivers’ tasks have become more challenging (competition for attention, higher frequency of important visual information) • Traffic densities higher • Traffic speeds faster • Vehicle forward lighting • Lamp and control technologies
Research Aims Overall Aims • To investigate the capacity of drivers to detect objects and pedestrians on the roadway edge under different dimming levels of LED streetlights
Specific Aims • Four lighting levels: three dimming levels (25%, 50%, 75%) and 100%; • Measure light levels at the eye and at roadway level; • Measure driver reaction times and pedestrian recognition distances in a young group of regular drivers with normal vision
Methods: Driving Circuit • Bitumen closed road circuit • Standard road signs and markings • Hills, bends, straights • 1.8 km loop of the circuit • No night-time lighting • Transport and Main Roads
Methods: LED Streetlights Six LED streetlights positioned at 50 m intervals along straight stretch • • • •
Mounting height 11 m, 2 m outreach LED Roadway Lighting (LRL) NXT 78M Controlled by wireless control and monitoring system (Lumen IQ) 25% (L1), 50% (L2), 75% (L3) and 100% (L4) of maximum output Primary Pedestrian
B
A
Mounted Streetlights
A Secondary Pedestrian
B
Methods: LED Streetlights • • • • • • •
LRL NXT 78M 72 (2 x 36) LEDs 158 W power factor 0.962 4000 K CRI: 73 GPS-based node control
Methods: Instrumented Vehicle Horizontal and vertical illuminance on car roof,
2015 automatic transmission Toyota Camry •
Low beam headlights
4 Konica Minolta illuminance sensors • • •
Vehicle roof Vehicle interior At the driver’s eye
In-vehicle GPS and response logger • • • •
Sampling at 4Hz Vehicle speed Driver reaction times Pedestrian recognition distances
GPS tracker Illuminance at the eye
vertical illuminance in the car
touchpad sensor
Methods: Participants 14 younger drivers • Mean age 34.2 ± 4.9 years, range 27-40 years • 7 males, 7 females
Licensed drivers: • 16.3 ± 5.0 years of driving experience • Weekly driving: 63 ± 48 kms (day); 27 ± 20 kms (night) • Self-reported driving ability: “average” or better • 91% reported “a little” or “no difficulty” for night driving
Attended 2 testing sessions • Vision Assessment • Driving Assessment
Methods: Vision Assessment Visual Acuity: • M = -0.14 ± 0.06 logMAR (1 line better than 6/6 or 20/20)
Contrast Sensitivity: • M = 1.97 ± 0.03 logCS
Visual Fields: • Within normal limits
Methods: Driving Assessment A
B
Light Barrier 2
Stationary Target - B
Moving Target
Light Barrier 1
Stationary Target - A
Primary Pedestrian
B
A
Mounted Streetlights
A Secondary Pedestrian
B
Methods: Driving Assessment Primary pedestrian located at either position A or B for repeated laps • Walking in place facing the driver • Grey street clothing
Moving target and a stationary target (alternating A and B position) • 1.8 km/hr (moderate walking speed) triggered by car passing 75 m away B
A
Light Barrier 2
Stationary Target - B
Moving Target
Light Barrier 1
Stationary Target - A
Methods: Experimental Design 9 laps: 1 practise, 8 test laps • 2 laps (A and B configurations) for each streetlight dimming level • •
25% (L1), 50% (L2), 75% (L3) and 100% (L4) of maximum output Order of laps and dimming levels was counterbalanced
• Participants instructed to drive at a comfortable speed and press a dash-mounted touch pad when they recognised either a person or moving target along the roadway B
A
Light Barrier 2
Stationary Target - B
Moving Target
Light Barrier 1
Stationary Target - A
Methods: Experimental Design 9 laps: 1 practise, 8 test laps • Reported all road signs (distractor task) • At end of each lap, rated level of agreement with the statement “The streetlighting on my last lap was effective in providing good driving visibility” 1. Strongly Agree; 2. Agree; 3. Neither Agree or Disagree; 4. Disagree; 5. Strongly Disagree Primary Pedestrian
B
A Mounted Streetlights
A Secondary Pedestrian
B
Results: Lighting Conditions
Results: Illuminance on Road Surface Measurements (in lux) made at 50% dimming (L2) Pole 4
Pole 5
¼
¼
19.3
13.0
8.2
4.8
3.7
3.6
5.1
8.1
13.2
20.5
18.0
12.3
8.1
4.8
3.8
3.7
5.1
7.8
12.1
17.8
Results: Road Luminance & Illuminance Dimming Level (% dimmed)
L1 (25%)
L2 (50%)
L3 (75%)
L4 (100%)
0.25
0.59
0.83
1.14
< V5 (0.35)
V4 (0.5)
V3 (0.75)
V2 (1.0)
% of L4 – road luminance
22%
52%
73%
100%
Average illuminance on the road surface (lux)
4.4*
10.3
14.5*
19.8*
-
6.8
-
-
Mean road luminance (cd/m2) Comparable road category, (luminance, cd/m2)
Illuminance uniformity (max:min)
*Projected values based on measurement at 50% and relative road luminance at other levels
Results: Illuminance at the Eye Illuminance at the eye along the lit section of the circuit Dimming Level
Mean eye illuminance with (SD) (lux)
L1
0.31 (0.13)
L2
0.55 (0.23)
L3
0.83 (0.34)
L4
0.95 (0.42)
Results: Moving Target Luminance Dimming Level
Moving target luminance (cd/m2)
Moving Target contrast ratio
L1
0.119
54%
L2
0.251
59%
L3
0.410
62%
L4
0.479
57%
Results: Primary Pedestrian Luminance Dimming Level
Pedestrian luminance (cd/m2)
Pedestrian contrast ratio
L1
0.065
10%
L2
0.168*
-
L3
0.235*
-
L4
0.322
21%
Primary Pedestrian
B
A
*Projected values based on measurements at L4 and relative road luminances at other levels A Secondary Pedestrian
B
Study Results & Discussion This research measured the effects of LED streetlight dimming on drivers’ reaction times and recognition ability • Streetlight dimming affected drivers’ capacity at night • • • •
Particularly for L1 (25%) and L2 (50%) relative to L4 (100%) Increased driver reaction times Shorter pedestrian recognition distances Drivers were aware to some extent of the streetlight dimming
Research informs the design of more extensive closed and open road studies of drivers’ behaviour and perceptions
Future Directions & Research Plan Timely to explore the human factor and safety aspects of streetlight dimming • • • •
Driver performance (driving reaction times, pedestrian visibility) Perception & preference (e.g. 3000K vs 4000K) Glare and visual comfort Ageing population impact on vision
Next experimental steps: • • • • •
Compare effects of LED dimming on driver performance with HPS Effects of colour temperature Effects of driver speed Extend the driver sample: older drivers and those with early visual impairment Open road studies
Acknowledgements Study funding support: Department of Transport & Main Roads Noel Peters, Greg O’Dea (Road Operations, Engineering & Technology Branch) Mt Cotton Driver Training Circuit
Study in-kind support: LED Roadway Lighting (LRL) Pecan Lighting Research team support Trent Carberry (Vision Sciences Group) Jason Koerper (QUT Photometric Laboratory) Volunteer drivers and other helpers
QUT Vision & Driving and Lighting Research Professor Joanne Wood; Dr Alexander Black •
Vision Science Research Group, QUT
Dr Gillian Isoardi; A/Professor Ian Cowling •
Lighting Research Group, QUT
Pilot study supported by: • Dept. Transport & Main Roads • LED Roadway Lighting
SLSC Roadmap 2016/17: on Research Few published studies worldwide have investigated the relationship between road lighting levels and crash rates at night Examples of Potential Research Projects: LED lighting application and road safety effectiveness Substantiate road safety benefits of LED (white light) Provide evidence for acceptable dimming conditions
‘Robust, credible research should underpin all policy measures’
Road Lighting and Safety Design principles of streetlighting have changed little since its inception around the 1930s, however: • Drivers’ tasks have become more challenging (competition for attention, higher frequency of important visual information) • Traffic densities higher • Traffic speeds faster • Vehicle forward lighting • Lamp and control technologies
Research Aims Overall Aims • To investigate the capacity of drivers to detect objects and pedestrians on the roadway edge under different dimming levels of LED streetlights
Specific Aims • Four lighting levels: three dimming levels (25%, 50%, 75%) and 100%; • Measure light levels at the eye and at roadway level; • Measure driver reaction times and pedestrian recognition distances in a young group of regular drivers with normal vision
Methods: Driving Circuit • Bitumen closed road circuit • Standard road signs and markings • Hills, bends, straights • 1.8 km loop of the circuit • No night-time lighting • Transport and Main Roads
Methods: LED Streetlights Six LED streetlights positioned at 50 m intervals along straight stretch • • • •
Mounting height 11 m, 2 m outreach LED Roadway Lighting (LRL) NXT 78M Controlled by wireless control and monitoring system (Lumen IQ) 25% (L1), 50% (L2), 75% (L3) and 100% (L4) of maximum output Primary Pedestrian
B
A
Mounted Streetlights
A Secondary Pedestrian
B
Methods: LED Streetlights • • • • • • •
LRL NXT 78M 72 (2 x 36) LEDs 158 W power factor 0.962 4000 K CRI: 73 GPS-based node control
Methods: Instrumented Vehicle Horizontal and vertical illuminance on car roof,
2015 automatic transmission Toyota Camry •
Low beam headlights
4 Konica Minolta illuminance sensors • • •
Vehicle roof Vehicle interior At the driver’s eye
In-vehicle GPS and response logger • • • •
Sampling at 4Hz Vehicle speed Driver reaction times Pedestrian recognition distances
GPS tracker Illuminance at the eye
vertical illuminance in the car
touchpad sensor
Methods: Participants 14 younger drivers • Mean age 34.2 ± 4.9 years, range 27-40 years • 7 males, 7 females
Licensed drivers: • 16.3 ± 5.0 years of driving experience • Weekly driving: 63 ± 48 kms (day); 27 ± 20 kms (night) • Self-reported driving ability: “average” or better • 91% reported “a little” or “no difficulty” for night driving
Attended 2 testing sessions • Vision Assessment • Driving Assessment
Methods: Vision Assessment Visual Acuity: • M = -0.14 ± 0.06 logMAR (1 line better than 6/6 or 20/20)
Contrast Sensitivity: • M = 1.97 ± 0.03 logCS
Visual Fields: • Within normal limits
Methods: Driving Assessment A
B
Light Barrier 2
Stationary Target - B
Moving Target
Light Barrier 1
Stationary Target - A
Primary Pedestrian
B
A
Mounted Streetlights
A Secondary Pedestrian
B
Methods: Driving Assessment Primary pedestrian located at either position A or B for repeated laps • Walking in place facing the driver • Grey street clothing
Moving target and a stationary target (alternating A and B position) • 1.8 km/hr (moderate walking speed) triggered by car passing 75 m away B
A
Light Barrier 2
Stationary Target - B
Moving Target
Light Barrier 1
Stationary Target - A
Methods: Experimental Design 9 laps: 1 practise, 8 test laps • 2 laps (A and B configurations) for each streetlight dimming level • •
25% (L1), 50% (L2), 75% (L3) and 100% (L4) of maximum output Order of laps and dimming levels was counterbalanced
• Participants instructed to drive at a comfortable speed and press a dash-mounted touch pad when they recognised either a person or moving target along the roadway B
A
Light Barrier 2
Stationary Target - B
Moving Target
Light Barrier 1
Stationary Target - A
Methods: Experimental Design 9 laps: 1 practise, 8 test laps • Reported all road signs (distractor task) • At end of each lap, rated level of agreement with the statement “The streetlighting on my last lap was effective in providing good driving visibility” 1. Strongly Agree; 2. Agree; 3. Neither Agree or Disagree; 4. Disagree; 5. Strongly Disagree Primary Pedestrian
B
A Mounted Streetlights
A Secondary Pedestrian
B
Results: Lighting Conditions
Results: Illuminance on Road Surface Measurements (in lux) made at 50% dimming (L2) Pole 4
Pole 5
¼
¼
19.3
13.0
8.2
4.8
3.7
3.6
5.1
8.1
13.2
20.5
18.0
12.3
8.1
4.8
3.8
3.7
5.1
7.8
12.1
17.8
Results: Road Luminance & Illuminance Dimming Level (% dimmed)
L1 (25%)
L2 (50%)
L3 (75%)
L4 (100%)
0.25
0.59
0.83
1.14
< V5 (0.35)
V4 (0.5)
V3 (0.75)
V2 (1.0)
% of L4 – road luminance
22%
52%
73%
100%
Average illuminance on the road surface (lux)
4.4*
10.3
14.5*
19.8*
-
6.8
-
-
Mean road luminance (cd/m2) Comparable road category, (luminance, cd/m2)
Illuminance uniformity (max:min)
*Projected values based on measurement at 50% and relative road luminance at other levels
Results: Illuminance at the Eye Illuminance at the eye along the lit section of the circuit Dimming Level
Mean eye illuminance with (SD) (lux)
L1
0.31 (0.13)
L2
0.55 (0.23)
L3
0.83 (0.34)
L4
0.95 (0.42)
Results: Moving Target Luminance Dimming Level
Moving target luminance (cd/m2)
Moving Target contrast ratio
L1
0.119
54%
L2
0.251
59%
L3
0.410
62%
L4
0.479
57%
Results: Primary Pedestrian Luminance Dimming Level
Pedestrian luminance (cd/m2)
Pedestrian contrast ratio
L1
0.065
10%
L2
0.168*
-
L3
0.235*
-
L4
0.322
21%
Primary Pedestrian
B
A
*Projected values based on measurements at L4 and relative road luminances at other levels A Secondary Pedestrian
B
Study Results & Discussion This research measured the effects of LED streetlight dimming on drivers’ reaction times and recognition ability • Streetlight dimming affected drivers’ capacity at night • • • •
Particularly for L1 (25%) and L2 (50%) relative to L4 (100%) Increased driver reaction times Shorter pedestrian recognition distances Drivers were aware to some extent of the streetlight dimming
Research informs the design of more extensive closed and open road studies of drivers’ behaviour and perceptions
Future Directions & Research Plan Timely to explore the human factor and safety aspects of streetlight dimming • • • •
Driver performance (driving reaction times, pedestrian visibility) Perception & preference (e.g. 3000K vs 4000K) Glare and visual comfort Ageing population impact on vision
Next experimental steps: • • • • •
Compare effects of LED dimming on driver performance with HPS Effects of colour temperature Effects of driver speed Extend the driver sample: older drivers and those with early visual impairment Open road studies
Acknowledgements Study funding support: Department of Transport & Main Roads Noel Peters, Greg O’Dea (Road Operations, Engineering & Technology Branch) Mt Cotton Driver Training Circuit
Study in-kind support: LED Roadway Lighting (LRL) Pecan Lighting Research team support Trent Carberry (Vision Sciences Group) Jason Koerper (QUT Photometric Laboratory) Volunteer drivers and other helpers
