It’s using the latest AI software for quality control, we’re hoping it’ll drive e-bike costs down and reliability up


E-bikes, washing machines, your toothbrush… it’s well known they all make weird sounds when they’re about to go wrong, and if we just listened hard enough we could figure out the fix too. It sounds like junk science, but now it turns out there really is something to be said for detecting faults by the noises they make. And Fazua is doing just that with its e-bike motors.

It’s more complicated than just cramming your ear as close to the motor as possible and holding your breath though – Fazua is using AI technology to listen for faults in drive units including the Ride 60 system, one of the best e-bike motors around. 

Fazuas lightweight e-bike motor, the Ride 60, out of a bike and shot side on revealing both the driveside and non-driveside

The Ride 60 motor is an impressive unit, it pushes out 60Nm of torque and weighs just 1.96kg

Its proper name is acoustic testing software, they’re calling it ‘Sounce’ and it’s said to analyse the sound patterns from the structure of the entire motor system, and identify problems in real time. It’s looking specifically for anomalies in the transmission and the bearings, as well as assembly errors, electronic faults and contaminated components. 

This has big implications for the cost of motors and eventually e-bikes themselves, because a large chunk of the initial outlay of an e-bike goes into the motor’s warranty. If brands like Fazua can reduce the failure rates of their motors before they’ve even gone wrong, we should see prices drop too. 

Santa Cruz Heckler SL

The Santa Cruz Heckler SL is one of our favourite lightweight e-bikes, in no small part to its use of the Ride 60 motor

Fazua is already saying there’s a 60% increase in cost-efficiency, and that’s just testing the motors as they come off the production line. Imagine if bike shops could use the service too, perhaps letting you know there’s a telltale sound of water inside your motor. It could potentially reduce the cost of maintaining and servicing some of the best electric mountain bikes, and cut the time they’re out of action being fixed too.

The AI Sounce software from third party firm MHP then goes on to learn from the errors it encounters and this should help it figure out how best to deal with problems, and to identify new sound patterns too. Or in engineering speak, “the deep learning algorithm will be trained in further error sources using existing data and additional data added by the engineers to increase its precision.”

It’s no surprise Fazua is using sound detection, given it’s part of the Porsche group and that the AI learning algorithm is already being used for cars. Skoda, separate to the group, even has an app for your phone that can recognise a car fault and diagnose it on the spot. Please Fazua, put this clever software into an app and let us look after our e-bikes better!

“The quality of our products is our top priority, Dr. Alexander Wünsch from Porsche eBike Performance said in a statement. “With the introduction of the acoustic testing software from MHP, we are taking the next step in quality assurance. The ability of AI to automatically identify technical problems in real time using sound and learn continuously will considerably optimize our production processes. It will also enable our highly qualified specialist personnel to focus on other work.”