Traffic Engineering - July 2019
UTSA engineers develop inexpensive smart stop sign to improve rural road safety
Engineers from the University of Texas are building and testing a new low-cost, self-powered system that will detect vehicles, improve the visibility of stop signs, and help prevent intersection deaths, particularly at rural locations.
Rural roads account for 70% of the USA’s byways and the location for 54% of all fatalities, according to the US Department of Transportation’s Federal Highway Administration. Without access to a power supply, they are more likely than other roads to lack signals and active traffic signage.
To improve driver safety, the UTSA’s College of Engineering, created a low-cost, self-powered intersection detection and warning system to alert rural motorists about potential dangers. The next-generation stop sign uses a multi-pixel passive infrared sensor that detects a vehicle as it approaches an intersection. Once the vehicle is within the sensing range, a signal beacon triggers the stop sign’s flashing system.
The off-roadway system can be installed on urban or rural roads completely independent of the utility power grid, because it is powered by small solar panels and functions in all weather conditions.
AITPM Fellow Graeme Pattison comments on this issue
There are frequent suggestions to install flashing lights at traffic safety trouble spots. If used judiciously these can be effective. If widespread, the benefits would be much reduced as the devices become normal background objects. Being responsive to vehicle speed, direction and category, these smart signs would give fewer false or unnecessary alerts. The red flashing light used in this project would not be acceptable in Australia as it would be deemed as a red traffic signal, so a yellow light would need to be used instead. Simple lights are not conspicuous in daylight unless surrounded by a large target board similar to that used on traffic signals. Stop signs, especially in rural areas, should be retro-reflective and hence highly visible at night and from a distance. There is a further problem that motorists at night avoid looking towards flashing lights due to dazzle and hence they are less likely to see the stop sign. The RMS operated CrashCam with 24 hour video recording at a number of rural intersections with high accident rates. At several sites a high percentage of motorists deliberately went through the stop signs at considerable speed when they believed there were no other vehicles nearby.
A video of the project team's work can be seen at: https://youtu.be/hMBLfV8BQ44
TU Graz developing ‘predictive’ pedestrian crossing signals for Vienna
Austria’s University of Technology is developing a system that recognises pedestrians’ intention, or not, to cross the road
It is well known that not all pedestrians wait for the green phase and cross the street when the lights are red, resulting in drivers having to stop even though nobody is there.
With this new camera-based system, if there is a pedestrian, the walk light stays lit long enough, for all the people it has detected, to safely get across. However, if the system sees that someone has approached the crossing but then left, it will cancel the walk-light request, allowing traffic to flow uninterrupted.
AITPM Fellow Graeme Pattison comments on this issue
In light traffic conditions traffic signals can be more time efficient if Walk signals can be cancelled where the pedestrian has already moved on. Detecting waiting pedestrians is a difficult task for camera systems although there are a number of suppliers in this market. Determining the direction pedestrians intend to go is even harder as pedestrians often stand well back from the kerb to avoid the slope of the kerb ramp and to be away from moving traffic. A reasonable accuracy should be proven before such a system is installed.
Australian university uses Google Street View to manage roadside infrastructure
Geospatial scientists from Australia’s RMIT University in Melbourne have developed a new program that uses artificial intelligence (AI) to monitor street signs needing replacement or repair by tapping into Google’s Street View images.
AITPM Fellow Alan Finlay comments:
I like the use of technology to identify the sign locations, but I have some reservations:
1. Google StreetView is not always up to date, so the rubbish truck cameras are probably better.
2. What if the sign is completely missing (i.e. just a stem in the ground)?
3. Will local authorities have the budget (or motivation) to fix up all the signs in poor condition?