misuse of scooters in the city is one of the main problems when it comes to zero-emission micro-mobility. Whether it is about vehicles abandoned in risky places, excessive speed or violations of the Highway Code, many local administrations are considering specific measures to regularize the use of scooters, and even the operators of sharing services are moving in this direction. .
Computer vision for scooters, how experimentation works
Swedish company Voi, for example, recently launched the world’s first large-scale “computer vision” trial, using artificial intelligence to understand when a scooter leaves the road to get onto a sidewalk or is parked incorrectly. . The technology belongs to Luna, a startup specializing in micro-mobility, and has the dual objective of protecting both those who use scooters and pedestrians.
The pilot project starts in Northampton in July, where Voi has an exclusive license from the UK government to test the scooter sharing service. However, the scooters equipped with the new technology were first tested on the streets of Stockholm, a trial that also served to identify areas of the city where it is difficult to take advantage of micro-mobility. The experimentation in Great Britain will instead develop in two phases, after Voi has installed cameras on a certain number of scooters.
In the first phase, a group of users will be identified who will test computer vision technology on the road to collect visual information in real time on the environment that the scooter is crossing, as well as detecting pedestrians on its path. The technology will also be able to detect the surface on which the scooter is moving, such as a bike path, a sidewalk or a road, and alert with an audible alarm to alert the user of the infringement. The AI will also be able to drive towards “virtual docks” where it is safe and allowed to park the scooter.
In the second phase of the trial, around 100 cameras will be installed on the fleet of scooters available in Northampton. In addition to the audible warning, this pilot phase will explore the potential to automatically slow scooters if inappropriate driving on pavements or in heavily pedestrianized areas is detected. In the same way, the Luna technology will allow you to park correctly using the camera as a sensor, a bit like what happens on cars. All vehicles will be tracked and the data collected to process a series of reports capable of highlighting any critical issues and problems. By 2022, Luna expects to be able to integrate its camera technology directly onto the scooters.
«With computer vision, scooters can be trained to see and recognize dangerous situations – said Fredrik Hjelm, co-founder and CEO of Voi Technology – This experimentation will establish new safety standards for this new form of transport. With over 60 million racing experience in Europe, we deeply understand the safety issues of scooters and are always looking for ways to do better. We are very proud to be the first operator to incorporate computer vision technology on a large scale for the benefit of our users, pedestrians and local authorities. “
“With this experimentation we can’t wait to demonstrate how scooters equipped with computer vision can make a difference in verifying users’ compliance with road rules – added Andrew Fleury, co-founder and CEO of Luna – We have noticed that cities all over the world are demanding technological solutions to counteract driving on pavements and it is fantastic to work with a safety-conscious operator like you in order to develop market-ready solutions. The value of smart city data that can be generated by vision scooters is only now starting to be understood and we are excited to explore these first use cases with you. “
