The Development of the Transport for London Progressive Safer System

Executive summary

The following report outlines the outcome of a series of activities that define potential options for a Transport for London Direct Vision Standard Progressive Safety System for HGVs. The Progressive Safety System (PSS) is a proposed revision of the Safer System (SS) that was defined as part of the original project to create a Direct Vision Standard for London in 2018.

Part 1 - Phase 1: Literature review

The literature review from the 2018 SS project (Knight, et al., 2018) has been updated and highlights significant improvements to the regulatory landscape for HGV safety, assuming GB implements all of the changes in the European General Safety Regulation.

• A UNECE Regulation on direct vision will improve the safety of all HGV categories through a demanding set of Direct Vision Requirements from 2029

• New regulations for Moving Off Information Systems (MOIS) and Blind spot information systems will be mandatory on HGVs from 2024, supporting the aims of the PSS.

• Since the development of the SS, new aftermarket blind spot information systems have entered the market which allow the differentiation between VRUs and objects which can be detected such as street furniture and parked cars.

• Germany has implemented a retrofit scheme for blind spot information systems that they call “turn assist” and financially supported the purchase of such technologies, if they complied with a relatively robust set of specification and testing requirements.

• Retrofit technologies for driver monitoring and the detection of fatigue and distraction have become available.

• The introduction of OEM technology such as AEB systems and the potential for EURONCAP to rate HGVs in future is likely to support the general improvement of HGV safety and the achievement of TfL’s Vision Zero strategy.

Part 2 – Phase 1: Existing solutions review

A review of existing solutions has demonstrated that there has been a marked improvement in the Blind spot information system technologies available, whilst also showing that low quality options exist.

• At the low end of the market, systems with identifiable quality and performance issues are being marketed as DVS compliant and sold on market places such as EBay for around £400.

• The SS offered relatively few specifications and did not require blind spot warnings to use a particular technology. It has become clear from the survey performed and contact with stakeholders that the passive Ultrasonic systems are the most commonly used solution but suffer operational limitations due to the inability to differentiate between VRUs and other objects. The range of these systems is often limited to 1m to reduce the number of false activations, which limits their effectiveness

• At the high end of the market RADAR systems and ultrasonic systems such as Brigade SideScan Predict are now able to differentiate between VRUs and objects such as street furniture and

parked cars. These systems are being fitted and testing byoperators such asConway andCEMEX. ¹

•Systems are entering themarketwhich are camera based and use ArtificialIntelligence/Machine Learning to identify VRUs.

Part 3 –Phase 1: Laboratory based testing of potential PSS elements–LU lead

Testing ofblind spot information systems based onbudget Ultrasonic sensors has highlighted anumber of short comings regarding reliabilityof detection, andvarietyof detection ranges being setat the pointof manufacture which is problematic in terms of variability in performance

Part 4 –Phase 1: Virtualtesting of the sensor and camera locations–LU lead

The virtual testingof Ultrasonic and RADARsystems has used specific accident scenarios to highlightthe benefits and shortcomings of these systems.

In particular passive ultrasonic systemswith limited lateral range havebeen shown in to ineffective ina number of key scenario stages for leftturn collisions with cyclists(themost common accidentscenario between HGVs and VRUs).

However the benefits of passive ultrasonic sensors to the frontof the vehicle,allowing VRUs to bedetected, have alsobeenshown.

Part 5 –Phase 2: Assessing equivalence of directvision and safe system technology–LU lead

This workemphasises thatdirectvision isvaluable tothe frontof the HGVcab and the area directlyadjacent to the passengerdoor, but is not effectivein identifying cyclists rearwards of the HGVcabduring the leftturn accident scenario.The benefitsofRADAR systems are shown.

The analysisof sensor types reinforces the shortcomings identified in part 4for passive ultrasonicsystems and highlights the benefitsof employing sensor systemsto the front ofthe vehicle to reducepedestrian casualtiesto the front of the vehicle.

Part 6 –Phase 2: Quantifying the state of the vehiclemarket

Interviewswith technical sales representatives from twomajor manufacturers selling vehicles in theUK have highlighted that 3starvehicles are possible and being purchased forthe regional and urbandelivery categories.The section highlights that there areoperational requirementswhich make itunlikely that 3 starvehicles will be specified for Long Haul and off road N3G HGVs. This part alsohighlights the improvements to direct vision found innew HGV cab designs produced by Scania andVolvo,which allow3 star constructionvehicles (butnotinN3Goff-road configuration).

¹The technology used does not determine the overall system performance, indeed, it is possible todesignaradarsystemthat does not differentiate between VRUs and otherobjects.Performancebased specificationsto allow developers and testingorganisations toestablish the true performanceof each system is needed to ensure that, for example, a radar system is in fact a good radar system.

Part 7 – Phase 2: Real world testing of sensor combinations – LU Lead

Real world testing of a range of systems has been performed which has informed the virtual testing performed in part 4. These have quantified the performance of systems with a range of quality from passive ultrasonics systems, active ultrasonic systems, RADAR systems and camera based AI systems.