What Van? Britain's biggest and best van buyers' guide

Over the past 12 months, VisionTrack, Europe’s leading provider of cloud-based video telematics has, using world leading algorithms, behaviour and video analytics continued to enhance and evolve its Autonomise.ai platform, providing this and associated services to its increasingly growing portfolio of customers. The product automates incident detection, data analysis and management processes, protecting drivers, mitigating risk and improving safety.

Although all dashcam solutions provide images of varying quality, it is how the information is analysed and interpreted that dictates its usefulness. Fleets of medium or large sizes don’t have the resources to view all automatically uploaded evidence of possible events, where the amount of videos could potentially number hundreds per day. Vans in particular, due to their weight especially when laden, can generate significant g-forces during driving, enough to prompt an alert from a dashcam and images to be uploaded to a fleet manager.

Launching NARA (Notification, Analysis and Risk Assessment) as an element of the existing platform, VisionTrack now provides a solution which acts instantly on images. Removing false positive events automatically, NARA quickly assesses incident footage using a computer vision model, then uploads the video to VisionTrack’s Autonomise.ai platform. Taking just eight seconds to validate the footage using advanced object recognition, if triggered, it will generate a priority alert. In addition, NARA calculates an Occupancy Safety Rating (OSR), immediately identifying if a driver or passenger requires medical assistance. 

An initial testing phase involved a client with a fleet of 900 vehicles which had been generating an average of 2,000 priority videos every week, potentially taking an individual over eight hours to manually review. Using NARA, these events were reduced to less than five per day. VisionTrack claims that as a result the system eliminates the risk of error, improves the speed of response to an incident, rapidly validating the welfare of the driver with the associated duty of care offered, reduces the number of disputed insurance claims and delivers an average of £2,000 in claims saving for each incident detected. With VisionTrack’s installed devices already recording an average of four million driver miles every day, which is equal to driving the UK’s road network 16 times over, this generates 63 million hours of video – the potential for NARA to cut this down to a manageable and relevant size for its users is considerable. 

NARA isn’t the only enhancement to the VisionTrack Ai platform over the past year. Vulnerable Road User (VRU) perception technology enables cameras to ‘understand’ human behaviour, allowing them to predict the actions of an individual, providing collision warnings that could lead to an incident being avoided.