18 September/October 2023 | E-Mobility Engineering In conversation | Marko Lehtimaki detecting individual grains of sand in the road. That gives us a huge advantage in writing smart software for keeping the rider in complete control over every aspect of their riding experience.” Data-driven EVs With the Verge TS, TS Pro and TS Ultra all now commercially available, the computing power and processors inside them have been factory-set, to undisclosed parameters but to the satisfaction of Lehtimaki and his team that all the software-defined feature enhancements they may want can be patched into the motorbikes remotely, once fully written and tested in-house. “We upgraded the computer hardware in the motorbike several times before scaling up manufacturing,” he says. “In Finnish engineering, you don’t ship something until you’re sure it’s perfect, and at 1200 Nm in the TS Ultra we’ve created a bike with the highest available torque in the world. But in software terms, we’re still far from using all the available power, and there’s much more to be explored. “That includes things like machine learning models for understanding terrain types, particular conditions and parameters of roads being crossed, how slippery the road is in real time; these have never been seen in motorcycles before. Couple that with data-driven decision-making and you can achieve automatic modes and adjustments in the motor itself, as well as in the HMI in terms of information, recommendations and graphics that can be presented to the rider.” AI research and its facets are nothing new to Lehtimaki, and he baulks at using terms such as AI, machine learning or anything else that could lead mainstream media to label his company’s product ‘the AI motorcycle’. However, he affirms that the motor’s biggest untapped potential lies in the execution of advanced algorithmic prediction engines that could maximise the usefulness of the direct-drive powertrain and other subsystems in the motorbike. “Machine vision and related data collection techniques will also result in real-time digital twinning of the motorcycle’s surroundings and decision engines for the driver’s controls, for greater safety and a better overall riding experience,” he says. “And while I’ve worked with those technologies before, doing it in a team that’s actually shipping motorbikes isn’t something anyone else has ever done until now. “It also means setting the safety bar for motorcycles higher than any previous bike OEM, so this isn’t just technology for technology’s sake; our users, their experience and safety are of paramount importance.” Experiencing AI While users’ quality of experience might sound an abstract concept, Verge separates it into three particular goals. One is the rider’s handling of the bike, a key part of which is moving the CoG to the bottom. “The software helps a lot as well though, not only the motor control, battery management and vehicle control, but also the interaction with the user and how they might be able to configure the bike’s riding modes for their own satisfaction regarding throttle response, regeneration, speed and torque,” Lehtimaki says. The second goal is safety, not just in terms of how the software might manage controls such as the automatic adjustments of the powertrain and controls on the user’s behalf, but also The direct-drive motor could enable millisecond-level control, with algorithmic prediction engines to assist the driver in hazardous road conditions or terrains The motorbikes are designed to behave as smart devices, with 5G connectivity and a detailed HMI for customising riding mode parameters and viewing stats
RkJQdWJsaXNoZXIy MjI2Mzk4