Tesla Model S in "Full Self-Driving" is a danger to motorcycle riders

A Tesla Model S car was in “Full Self-Driving” mode when it hit and killed a 28-year-old motorcyclist in the Seattle area this last April 2024, making it the first fatal motorcycle accident involving Tesla's autonomous vehicle tech.

The 56-year-old driver was arrested on suspicion of vehicular homicide based on his admission that he was looking at his cell phone while using the driver assistant feature, police said in a statement.

AVs (Autonomous Vehicles) don't not look like they will increase safety for motorcycle riders anytime soon: "Speaking in an interview with the Tesla Owners of Silicon Valley club last weekend, Musk said a future vehicle will be like a “tiny mobile lounge” where drivers will be able to watch movies, play video games, work and even drink and sleep." OMG, I can't imagine anything worse for motorcycle transportation safety.

OK, if we compare how many human drivers have killed motorcycle riders to Teslas, there's no contest human drivers are a greater danger today, but as AVs transform our transportation systems, the problem becomes more serious with unlimited consequences for motorcycle transportation.

The problem with autonomous vehicles not recognizing motorcycles is that they use a camera system to "see" or to accurately measure safe distances. With camera-only systems it's just that: the camera only! Lighting, weather conditions, vibrations, all affect the camera which is not self-maintaining. I believe the future safety of motorcycle riders requires autonomous systems to combine real motorcycle data to camera systems to accurately detect and predict motorcycle movement: the kind of data the autonomous industry doesn't have today.

Motorcycle data collected in real-time from sensors in the smart-phone can be used to augment AV perception systems in obvious ways. The real-time data captures how the motorcycle moves, which is different from how a car moves. This type of sensor data can be used to train machine learning algorithms to better detect and classify motorcycles on the road. It can also provide information about the size, shape, speed, and movement patterns of motorcycles, while maintaining anonymity of the rider. This can help AVs recognize motorcycles quicker and more accurately, preventing them from crashing into us.

What I'm saying is that by combining and adding the sensor data ESR provides to detect motorcycle crashes to camera tech like lidar, AVs can create a more accurate digital model of the motorcycle which makes it possible to "see" the motorcycle using its "data ride-print". As a result drastically improving the overall performance of the perception system.

Yes, predicting the movements of motorcycles is challenging because we are small and agile. Basically, we're as fast as a car, but as visible as a pedestrian! Maintaining our right to ride in a self-driving car future is something we can't take for granted.

Motorcyclists are vulnerable road users; we need protection from Autonomous Vehicles and our transportation safety must be a priority for urban planners and ministries of transportation world-wide.

Read the source article on CNN here.