The UMV was developed as part of the RAI’s \u201cAthletic Intelligence\u201d program.\u00a0 The aim is to teach machines to move with agility and precision – skills that are otherwise reserved for human athletes.\u00a0 The system relies on reinforcement learning, a method of machine learning in which the robot improves its skills through repeated testing.<\/p>\n
\u00a0<\/p>\n
The vehicle is based on a children’s balance bike – more precisely a 500 pound (approx. 585 euros \/ 635 USD) Specialized Hotwalk made of carbon.\u00a0 An electric motor has been integrated into the frame, which drives the rear wheel via a thin belt.\u00a0 It is controlled by a robust mechanism equipped with sensors and processors.<\/p>\n
A special feature is the adjustable arm structure, which enables the UMV to perform bunny hops.\u00a0 The arm moves upwards at lightning speed, lifting the entire vehicle and ensuring controlled jumps.\u00a0 Landing is also precise: precisely controlled throttle pulses often cause the bike to touch down with the rear wheel.<\/p>\n
\u00a0<\/p>\n
It remains unclear why the RAI launched this project.\u00a0 The institute is pursuing the goal of overcoming new challenges in robotics and AI – the extent to which stunt riding plays a role in this is not immediately clear.<\/p>\n
In a published video, the RAI shows how the UMV drives through an office, avoids obstacles, masters jumps and comes to a standstill with a track stand.\u00a0 The high level of precision and control demonstrated by the robot underlines the progress made in robotics.<\/p>\n
\u00a0<\/p>\n
The UMV is not the first robot on two wheels.\u00a0 Yamaha has already developed Motobot, a machine that can ride a motorcycle autonomously.\u00a0 But the UMV goes beyond this and proves that machines can not only ride, but also perform complex stunts.<\/p>\n
This development raises questions: Is it purely for research or is there more to it?\u00a0 Autonomous motorcycles could be used for various purposes in the future – be it for transportation solutions, rescue missions or possibly even military applications.<\/p>\n
The technology is at an early stage, but the UMV project shows how quickly autonomous systems are developing.\u00a0 What is still an experimental stunt robot today could soon open up new possibilities for self-driving two-wheelers.<\/p>\n
[amazon bestseller=”Terminator Actionfigur” items=”3″]<\/p>\n\n\n\n