In August, 2022, the Connected Motorcycle Consortium released a white paper that examines how well Advanced Driver Assistance Systems (ADAS) detect a range of different powered two-wheelers (PTW)—also known as bikes, to you and me. The paper analyzed the current state of detection, as well as upcoming European New Car Assessment Program (Euro NCAP) requirements.
Most importantly of all, after taking a thorough dive into existing data, the CMC’s study offered recommendations to help ADAS systems “see” motorbikes better out on the road. The situation could clearly be better—and two riders who both died in suspected ADAS-related crashes involving Teslas rear-ending them—are just the latest proof.
So, what are some of the reasons that ADAS has trouble registering motorcycles—and more importantly, what can be done about it? After all, it’s in everyone’s best interest to solve this problem. Riders obviously don’t want to get hit, and most drivers (probably) don’t want to hit us.
As ADAS systems have become more widely available, a big problem with PTW detection is that it hasn’t always been analyzed as its own thing. When ADAS was in its infancy, analysts took a look at pedestrians, bicyclists, and other four-wheeled traffic on the road (trucks, cars, and etc.) Somewhere along the line, an assumption was made that if systems accounted for pedestrians and cyclists, then motorbike riders would be covered as well.
If you’re spluttering loudly and/or making a wide array of irritated hand gestures because of that paragraph, we don’t blame you. As riders, we know very well that the way motorbikes behave is not usually the same as either pedestrians or cyclists. For one extremely important point, most motorbikes—be they scooters or massive touring bikes—are much faster than pedestrians or cyclists.
Although motorbike vehicle size is significantly smaller than cars and trucks, all three types of vehicles often share roadways together. The conundrum seems to be that motorcycles can go car and truck speeds (or faster), yet simultaneously also have smaller detectability for ADAS systems, which is more akin to pedestrians and/or cyclists.
One CMC finding actually noted that in some cases, bicycles were more easily detected than motorbikes—probably due to the comparative flatness of their profiles, and the aerodynamic nature of many motorbike profiles.
From 2023, the Euro NCAP ratings system will have specific test criteria for ADAS systems as regards motorcycle detection—which is at least somewhat reassuring. It’s worth noting here, as CMC does in its white paper, that the Euro NCAP ratings system is constantly undergoing revisions.
When an auto maker says that its vehicle has an “X-star Euro NCAP rating,” that’s why it’s crucial to know what year that rating was achieved. Because criteria are constantly changing, the year is the best way to know what factors were considered at the time of the grade.
The CMC’s paper, incidentally, mainly discussed systems involving radar and/or LIDAR as components in hazard detection in ADAS. In other words, camera-only systems—such as the ones allegedly implicated in those two recent Tesla-on-motorcycle deaths—were not discussed.
The CMC suggests that motorcycle OEMs consider adding radar reflectors at key points on their bikes, in order to help ADAS systems better see those motorbikes out on the road. It additionally suggests that greater attention needs to be paid to precisely finding the right signal-to-noise ratios to detect motorcycles and their unique lane usage patterns on the road.
Since motorbikes may utilize multiple positions within a single lane of traffic, it seems that many ADAS systems don’t always see a bike unless it’s directly in front of them. If a bike is at the hard left or hard right side of a lane, or if it’s switching lanes, then existing systems may not always detect the bikes and respond appropriately.
During some tests, ADAS systems did alert their drivers that a hazard was detected—but in the case of radar-adaptive cruise control scenarios, did not always automatically brake to avoid a collision as soon as they should have. Manual driver intervention was necessary to avoid a crash in these scenarios.
As drivers become more comfortable with ADAS systems, and also build expectations that such systems will be able to respond appropriately to road hazards around them, systems that don’t or can’t account for normal motorbike behaviors (such as lane positioning that isn’t in the direct center of the lane) endanger riders. Numerous specialists in the field are working to advance these systems—but radar reflectors are already prevalent in the boat world. Perhaps they need to become more prevalent in the bike world as well.