The Motoroid 2, a revolutionary self-riding electric motorcycle from Yamaha, does away with conventional controls like handlebars. This version improves upon the Motoroid concept from 2017, which has been developed into a working prototype. By using gyroscopes and AI-powered image recognition technologies, the bike can ride itself without a rider while maintaining balance and navigating highways.
“Motoroid 2 is a vehicle for personal mobility that can recognise its owner, get up off its kickstand and move alongside its rider,” the company said.
“[It] has a distinctly lifelike feel when somebody is riding on its back and has a presence more like a lifetime companion.”
“Motorcycles will never ride autonomously; it doesn’t make sense,” said Dr Markus Schramm, head, BMW Motorrad. And rightly so. Motorcycles are inherently unstable vehicles with a shifting centre of gravity.
Motorcycling is bodily-kinesthetic. Depending on the type of motorcycle, the rider’s body position and technique need to change to get a handle on the machine. And with experience, motorcyclists develop muscle memory and intuition to make split-second decisions.
Second, motorcycles have low G-force as opposed to aircraft or cars. In the case of the former, we don’t have a lot of formalised literature or data on the bike-rider dynamic in the face of G force. To wit, G’s on a motorcycle is a bit more complicated.
Third, motorcycles are weight-sensitive. Considering how energy-inefficient today’s AI chips are compared to human brains, you’d have to pack huge batteries to endow the chip with enough “intuition” to ride a motorcycle, leading to a massive weight disadvantage.
Back in 2018, BMW developed a self-driving motorcycle with the ability to self-balance – accelerate, lean, and stop. However, it requires a human operator. The bike takes commands from the human operator via the antenna at the back. At the time, BMW said it had no plans to commercialize the project.
Key points:
1 Motorcycles require constant shifting body positioning and technique based on conditions, something current AI technology cannot adequately replicate. Human riders develop muscle memory and intuition for split-second reflexes.
2 Motorcycles experience more variable G-forces than cars or planes. There is insufficient data and models capturing the nuances of bike-rider dynamics in these situations for an AI system to safely control a motorcycle.
3 Weight sensitivity limits how much battery power can be added to enable autonomous functionality without compromising motorcycle handling and efficiency. Existing AI systems are far too energy inefficient compared to human cognition.
While some limited self-balancing functionality is possible, as BMW demonstrated on a test model, removing human operational control poses too many unsolved stability, dynamics, and energy efficiency challenges. Ultimately motorcycling intrinsically depends on human bodily movement, instincts, and reactions. An autonomous motorcycle thus remains implausible. Companies would be better served focusing innovation on rider-assistance features rather than eliminating the human element central to motorcycling.
Sources:
- https://www.independent.co.uk/tech/yamaha-driverless-motorbike-motoroid-2-b2427927.html
- https://evrevolutionhub.com/yamahas-electric-marvel-the-self-balancing
- quillbot.ai
- https://analyticsindiamag.com/self-driving-motorcycles-slip-a-gear/
Here are couple of videos concerning topic ‘motorcycles and ai’ if you are interested:)
https://spectrum.ieee.org/ghostrider-the-self-driving-motorbike-that-launched-anthony-levandowski
https://www.wired.com/2016/08/get-know-aboard-self-driving-motorcycle/
The Yamaha Motoroid 2 sounds cool with its ability to recognize its owner and move alongside them. Even though the idea is interesting, the problems with stability and energy efficiency make me a bit hesitant. Maybe, for now, if I wanted to buy a vehicle, I’d prefer a motorcycle with smart features to help the rider rather than a completely self-riding one.
In my opinion, while the Motoroid 2 showcases some impressive AI and robotics engineering, the idea of a fully self-riding motorcycle seems implausible and removes the inherent joy of riding. Motorcycling has always required riders to constantly adjust their bodies, develop memory reflexes, and rely on their instincts when on the road. The delicate weight shifts, balancing maneuvers, and split-second reactions required to control two wheels on a variety of roads cannot be replicated by a computer.
I really liked the conception of autonomous motorcycles presented by BMW, but overall the motorcycle itself is supposed to be driven by a human and riding a motorcycle brings the adrenaline rush and extreme emotions.