Shell Global Solutions International B.V. has commissioned SWOV Institute for Road Safety Research to create a combined inventory and qualitative comparison of commercially available Blind Spot Monitoring Systems (BSMSs) for buses, heavy goods vehicles and other large mobile equipment; out of an interest to equip their fleet with such systems. In addition, Shell requested an overview of measures other than BSMSs to mitigate blind spot related risks.
Research method
In order to perform a comprehensive inventory, lists of search terms and qualitative evaluation criteria were defined on the basis of literature and regulations. The resulting search terms were combined into a series of search queries which were presented to the Google search engine (cf. the project proposal). Eighty-three systems were initially found. For these systems, documentation was collected and/or requested directly from manufacturers. After further inquiries and evaluation of obtained documentation, 60 systems by 33 manufacturers were ultimately deemed suitable for further consideration. We described the properties of each system following a template of qualitative evaluation criteria, and aggregated the data in a database.
The aggregated data were evaluated in terms of completeness and general data quality. Although the Operational Design Domain of systems was apparent, considerable differences were observed between manufacturers and systems in terms of available data on System Implementation. Hardly any data were available on detection performance (e.g., false alarm rate, proportion of missed targets) and usability, making comparisons of Scientific Validity impossible. Quotes were obtained for 25 systems. Ultimately, the proportion of available observations permitted a ranking scheme for 7 system properties, namely: whether the system was (1) suitable for detection of vulnerable road users, whether it was (2) triggered (i.e., operational specifically under a predefined set of circumstances), whether it was (3) compliant with specific regulations, (4) unit costs, and, more specifically regarding the sensor(s), the (5) Ingress Protection rating, the (6) size of the Field of View, and (7) the range covered by the system.
Each system was ranked on the basis of each of these properties, in so far as the data were present. When data for a given property were not available, the system was ranked last for that particular property. The ranks were summed to obtain a total score (i.e., a ‘sum of ranks’) and systems were then ordered from best to worst based on this ranking.
Considerations
Apart from the method to rank systems included in the inventory, there are additional considerations to either favour certain systems, or to preclude them from application. These considerations are:
- Given an existing fleet, it is not feasible to replace all vehicles. This means that OEM systems that are only available for new vehicles were not selected for ranking.
- For systems that warn a driver but do not intervene, responsibility ultimately remains with the driver. Therefore, intervening systems may be preferred. However, it was found that these do not exist as retrofit solutions, presumably because adequate performance cannot be guaranteed when systems are installed by 3rd parties.
- Finally, a system that monitors the entire surround rather than a specific portion is likely to provide a higher degree of safety. Therefore, integrated solutions or single manufacturers that provide solutions to monitor all directions are preferable.
Top ranking BSMSs
Ideally, one makes use of detection performance (e.g., false alarm rate, proportion of missed targets) to rank systems in terms of crash prevention performance. However, data on detection performance were rarely provided. Therefore, sensor specifications (e.g., field of view, range) were used as proxy, in addition to several other properties described above. Taking the above considerations into account, the following systems came out on top:
- Overall, the best ranked system is the Autel Blind Spot Assist (Appendix E.1.1). However, this system is suitable only to monitor objects in the vehicle’s lateral blind spots.
- The eXia Active Sideguard system (Appendix E.14.1) is among the top three for all viewing directions. However, this system is rather exotic, in that the sensor technology (monitoring the electric field) is not used by any other manufacturer. We do not know how well this system works in practice (scientific validity).
- The next-best multi-directional solution using more common radar technology is the Sensata PreView Sentry (Appendix E.27.2). Another advantage of choosing this particular manufacturer could be the availability of variations of this system (see Appendix E.27)
- The best ranking multi-directional camera system is the oToGuard system by oToBrite. This is a single integrated solution to monitor all directions, which also features an in-cabin camera to monitor the driver, and which is compliant with multiple UNECE regulations (Appendix E.22.2).
Inventory of non-BSMS related measures
SWOV reports and fact sheets were used for an inventory of non-BSMS related measures. These sources mainly focus on blind spots directly to the right and in front of the cabin of HGVs, in interaction with cyclists and to a lesser extent pedestrians. Three types of measures were derived: 1) measures to support the driver (e.g., education, safety culture, blind spot mirrors, cabin design, road-side mirrors), 2) measures to separate HGVs and other road users in time and space (e.g., system changes, time windows, route choice, intersection design), and 3) measures to support other road users (e.g., signage, education). The consulted literature did not provide scientific evidence on the effectiveness of these measures for traffic safety. The transferability of these measures to other blind spots is discussed. Driver education and safety culture of companies relating to blind spot situations may help in all blind spot scenarios. Strategically placed blind spot mirrors at dedicated areas for collecting or delivering goods can help reduce the blind spot behind the vehicle. When working vehicles are manufactured or bought, it is wise to consider optimising the direct view around the vehicle as much as is possible. Working with relatively safe time windows to drive with HVGs will reduce all blind spot problems to a degree. Safer routes and intersection measures are likely restricted to interactions with vulnerable road users. The use of special time windows and safe routes can be applied to the private premises of companies (e.g., ban driving with HGVs where and when co-workers walk during lunchtime).
Conclusions
In general, it was found that systems using ultrasonic technology were suitable mostly for short-range, low-speed manoeuvring, whereas ultrasonic and camera systems are also suitable for longer-range and motorway use. When taking into account that a modular system suitable to monitor all directions may offer the most comprehensive solution, we find that a particular system using unique technology that monitors the electric field ranks best. However, somewhat lower, but nonetheless comparable, rankings were found for a more traditional camera system and radar system. Ultimately, it should be noted that although information could typically be retrieved on system Operational Design Domain and Implementation, reliable indications of their actual detection performance were only available for a single system, and performance for this system was limited to approximately 50% detections. Ecologically valid empirical studies on detection performance may help improve the discrimination of BSMSs.