Work zones signage control through Computer Vision to improve Road Safety

Initiative details

Work zones on roads are especially dangerous areas for construction workers, drivers and motorcyclists, and are difficult to interpret by driver assistance systems (ADAS). Pedestrians, cyclists and motorcyclists Compliance with the procedures and correct maintenance of both vertical and horizontal signage allows the risks in these areas to be minimized.
An autonomous road inspection, using systems based on Computer Vision and sensors on board vehicles, allows frequent monitoring of vertical signs and objective analysis of the quality of road markings, facilitating instant corrective actions and the storage of evidence of the status of that area after each check.
Institutions such as Interbiak, Provincial Council of Bizkaia and Provincial Council of Gipuzkoa already use the ASIMOB Autonomous Road Inspector for frequent monitoring of their work areas.
This monitoring allows the risk prevention teams at the works to receive notices every time, in each check, a signal changes with respect to the previous check or fails to comply with the established procedure. In this way, immediate corrections are made and the work is maintained in optimal conditions. The solution saves evidence of all checks.

Initiative date

Who was/is your target audience?

Public authorities

Topic

Improve vehicles and infrastructure

Organisation details

ASIMOB (Advanced Services In MOBility SL)
Enterprise
Spain
Bilbao

Contact name

Ibon Arechalde

Telephone number

+34657652776

Project activities

If you work together with external partners, list the most important partners and briefly describe their role.

The analysis through Computer Vision is applied on the cloud in servers provided by Google, Microsoft or AWS. This is the only external partner during the service.

Please describe the project activities you carried/are carrying out and the time period over which these were implemented.

The service is applied on work zones in the roads and streets managed by public authorities. It is applied all along the work, using the vehicles that work in that place to monitor automatically all the traffic signs. It has been applied in several works in the province of Bizkaia, in Spain, during more than 9 months in each zone, and also has been applied in Canada.

In order to understand better it, here is a video of Spanish National TV explaining the Autonomous Road Inspector in Work Zones:
https://vimeo.com/934829224/0a9a499092?share=copy

And here a video of challenge to detect errors in signage in work zones, to understand the improvement against a manual supervision::
https://vimeo.com/831269342/6c7070a853?share=copy

In terms of implementation, what worked well and what challenges did you need to overcome?

Good points:
The notifications are received everyday, according to the procedures established with the public entity. Evidence are generated everyday, so the incidences can be checked accessing to the video and also analysis can be applied later, to improve the procedures in these risky zones.
Challenge:
The service is not applied systematically in all the work zones, so some places are monitored and others not.
The information obtained thanks to this Autonomous Road Inspector should be integrated in navigators like Waze, Google Maps and Apple Car, but right now it is difficult to share our information with them due to several factors.

Evaluation

Please summarise how you have evaluated the initiative’s impact (e.g. social media reach, survey, feedback forms, statistics).

An example to understand the impact: only in 5 km of work zone, in Spain, during the 9 months of a work, more than 280 incidences were detected in the traffic signage.
In Spain, one of the safest countries in the world, according to data from the National Registry of Traffic Accident Victims, in 2021, 158 traffic accidents took place related to maintenance and conservation works on the roads.
Currently, the signage in the work zones is not monitored frequently or it is done manually. Nor is evidence kept of the way in which a work has been marked. This absence of monitoring makes it impossible to analyse the causes of accidents and improve safety measures in them.
The frequent application of the Autonomous Road Inspector in work areas makes it possible to address this serious problem systematically, reduce risk situations and analyze the causes of accidents that occur in these areas. All this, at a very low cost, thanks to the automation of processes through Artificial Intelligence.

The places where it is being used are already safer areas

What has been the effect of the activities?

Until now, the results obtained in the service and the notifications are received only by the public entity that is managing the monitored road or street. The information is used to improve the signage in the moment and to generate evidences of the real status of the infrastructure. This will ensure that the procedures are complied in work zones.
As explained before, in later stages, this information could be shared with navigators (Waze, Google Maps and Apple Car) to give this information to the cityzens.
When the service is applied systematically in all work areas, it will be possible to see how much the number of accidents related to maintenance and conservation works on the roads can be reduced.

Please briefly explain why your initiative is a good example of improving road safety.

Monitoring the infrastructures (not only work zones) to detect the incidences and the defects of the assets is a good practice that could reduce the 30% of accidents that are produces or worsened due to the bad status of the infrastructures.
This figure (30%) is explained in several studies in Denmark, UK and Spain, but it is rare to find actions that focus on mitigating these factors.
Spanish Road Association and other internacional associations (as IRF) are promoting the use of technology to improve the control of the status of the infrastructure. The use of technologies such as Artificial Vision greatly simplifies its application, although there is great reluctance to change in the sector.
The Autonmous Road Inspector has been already used in 9 countries: Spain, Belgium, The Netherlands, Brazil, Canada, Serbia, Romania, Saudi Arabia and Israel. The specific use case to monitor Work Zones has been already used in Spain and Canada.
It is replicable in any country and scalable by design.

How have you shared information about your project and its results?

We disseminate the information in the sectorial congresses: IRF International Congress, Intertraffic, Spanish Road Week...
Also, in TV: here is a video of Spanish National TV explaining the Autonomous Road Inspector in Work Zones:
https://vimeo.com/934829224/0a9a499092?share=copy

Supporting materials