A few weeks ago Alex Tilkin published an article summarizing some of the challenges that Ride Vision is dealing with in order to bring CAT™ – Collision Aversion Technology to motorbikes.
Indeed, the problem that Ride Vision solves is unique and intriguing. Preventing and reducing motorbike accidents demands peculiar approach, specific to the domain of 2-wheelers.
Here, I’d like to show one more challenge that the 2-wheelers impose on the proposed solution. The Threat Analysis challenge. Threat Analysis model at Ride Vision is responsible for understanding the actual threats/hazards imposed on the motorcycles and raise the alerts accordingly.
To demonstrate the challenge, please see below a short footage, recorded by one of the Ride Vision’s systems, representing the dynamic nature of a 2-wheeler maneuvers.
This is quite normal and standard behavior of a motorbike in traffic, but it emphasizes the difference between the cars and the 2-wheelers. The above footage doesn’t contain dangerous situations and hence Ride Vision’s Threat Analysis model, which was built ground up to deal with the unique behaviours of motorcycles, takes into the account much more than the classical methods of analysis, detection and tracking of a motorbike’s surrounding.
Ride Vision’s Threat Analysis learns and deducts patterns of motorcycle behaviors on the road and constantly blends and fuses all the data points to infer instantly whether any threat can emerge.
In this case the Threat Analysis model won’t raise an alert. This exactly why the power of the Predictive Vision allows Ride Vision to bring its solution to any motorbike. Even a bit extreme examples (like below) can be inferred; Patterns can be learnt and predicted by Ride Vision’s Predictive Vision.