TfL’s AI Cameras branded ‘Subpar’

Transport for London (TfL) has been criticised for allowing the use of a ‘subpar’ AI camera system to meet the DVS requirements, according to a transport industry expert. TfL has indicated that installers can ‘self-certify’ their BSIS and MOIS systems on the premise that they adhere to the necessary specifications to comply with the October regulations. However, this approach has led to many misunderstandings among system manufacturers and misinformation provided to operators.

While AI cameras may seem more economical, they could incur significant costs when putting profit before safety. Independent testing conducted by Loughborough University revealed that an AI Camera failed to meet the new specifications, primarily due to its susceptibility to adverse road and weather conditions, and poor lighting. This limits the camera’s ability to achieve the requisite 100% performance level and effectively detect Vulnerable Road Users (VRUs).

Based on these findings, TfL must revoke permits issued where AI cameras have been self-certified to meet the new PSS, unless evidence to the contrary can be provided. When these systems fall short of the required performance level, they not only endanger lives but also place operators in a precarious position in the event of serious injury or fatality. SMUK’s Chairman Steve MacDonald says, “I am utterly gobsmacked that some of the subpar AI cameras made it through as an option for a blind spot.

It is not fit for purpose, and we at SMUK will not fit them.” SMUK’s due diligence process ensures it only fits equipment suited for all weather and lighting conditions that HGV operators may encounter. Brigade Electronics has unveiled Radar Predict, a cutting-edge AI-powered Blind Spot Information System (BSIS) that protects cyclists from collisions with HGVs.

Unlike AI camera systems, Radar Predict has undergone rigorous testing by SMUK, a leading safety solutions provider. Radar Predict utilises AI to analyse data like vehicle speed, direction, and nearby cyclists. It alerts drivers of potential collisions through a dual-radar unit fitted to the vehicle’s nearside, providing complete side coverage, including the trailer.

The system issues differentiated visual and audible alerts based on the time to collision, with escalating warnings for imminent risks. It’s not impacted by low level light conditions, dirt or weather. Radar Predict has several key features, including 180-degree side detection up to 46m at speeds below 20mph, compatibility with rigid and articulated vehicles, and a trailer discovery mode to prevent false alerts caused by articulation.

Brigade Electronics’ senior product manager, Thierry Bourgeay, says, “Radar Predict addresses blind spot issues and reduces false alerts, automatically activating during turns without requiring indicator selection.”

The system aims to enhance driver visibility and reduce the risk of collisions, contributing to improved road safety for all users.