16-04-2021 | By Sam Brown
Recently, The Highland Council has secured a £180,000 budget for deploying an AI traffic detection system that will provide data on traffic volume and type of traffic. But, is there a cause for privacy concerns, or will the system be a benefit for future generations?
The Highland Council (Inverness) was recently granted a £180,000 budget from the Scottish Parliament to deploy AI traffic monitoring systems to understand road usage better. 34 such sensor systems have been deployed at key locations, including notable roads and junctions. The same system is also expected to be deployed in areas that have created temporary cycle and walking paths under the Spaces for People initiative.
By better understanding the traffic type and volume that roads service, it is believed that future planning of those roads can be better done. For example, the traffic volume of a stretch of road can be easily determined with simple counters. Still, if the type of traffic is understood (cars, vans, lorries etc.), then specific decisions can be taken such as truck-only lanes, increased cycle path size, and congestion charge zones.
The funding of the project covers the cost of the sensors and the software and maintenance for the next five years. The need for a long time scale most likely comes from the fact that AI works better the more data it is fed, and since traffic can vary seasonally, having multiple years of data provides a solid dataset for the AI to work off.
While the council runs the project, the underlying technology is coming from a company called Vivacity Labs. Vivacity Labs is a company that specialises in AI technology surrounding traffic systems and sensors that can gather detailed and anonymous data 24/7. Their technology also extends beyond simple traffic monitoring; traffic flow data can be tied to Smart Junctions, helping traffic networks better organise traffic flow.
Their system can identify up to 9 different forms of transportation including van, truck, car, and bicycle, while also gathering metadata including speed, vehicle path, mean journey time, and social distancing. However, with privacy becoming an increasing concern, Vivacity Labs states that they operate with privacy at the heart of their work. This is expressed in blurred data and preprocessing that helps prevent private data from being accessible outside of devices. Furthermore, the sensors developed by Vivacity Labs do not collect personal data, and this inability extends to customers of their technology.
Vivacity Labs have clearly stated that their devices do not collect private data, nor will they ever allow customers (i.e. Highland Council) to do so. However, while the initial project may prevent private data collection (such as reg plate and face ID), future systems may not follow the same standard.
The emphasis on privacy on the initial project provides a safe narrative to customers with its genuine safety concerns. Still, if the project is implemented widespread, cheaper systems may be favoured to save money. Generally speaking, cheaper systems can cut corners or use off-the-shelf hardware and allow for the collection of private data.
The ability to collect private information such as journey routes, faces, and reg plates may also provide a source of temptation for the authorities who typically desire such technology. As such, the success of the initial project could encourage the erosion of privacy down the line.
However, if the use of privacy-keeping sensors is used, the use of AI with traffic systems could boost both commuters and traffic budgets. Furthermore, smart junctions powered by traffic AI could provide safer roads, which would reduce the number of road fatalities all while helping to reduce the overall emissions from vehicles, as such creating a safer environment.