AI-pow­ered dri­ver assis­tance sys­tem for micro­mo­bil­i­ty vehicles

As part of an inno­v­a­tive devel­op­ment project, a dri­ver assis­tance sys­tem specif­i­cal­ly tai­lored to e‑bikes, e‑scooters, and oth­er elec­tri­cal­ly pow­ered micro­mo­bil­i­ty vehi­cles was developed—with the clear goal of sig­nif­i­cant­ly improv­ing safe­ty in urban areas.

At its tech­no­log­i­cal core is a high-per­for­mance cam­era sys­tem designed for object detec­tion and track­ing. Com­bined with sig­nal pro­cess­ing opti­mized specif­i­cal­ly for micro­mo­bil­i­ty and intel­li­gent algo­rithms, the sys­tem enables fea­tures such as lane-keep­ing detec­tion and col­li­sion detec­tion with oth­er road users or oth­er hazards.

The par­tic­u­lar chal­lenge was to imple­ment object recog­ni­tion in a resource-effi­cient man­ner and in real time on com­pact hard­ware. This involves the use of both tra­di­tion­al image pro­cess­ing meth­ods and AI-based approach­es, tai­lored to ener­gy-effi­cient embed­ded sys­tems. Our strength lies in opti­miz­ing com­plex algo­rithms for reli­able oper­a­tion on resource-con­strained hard­ware, as well as in tak­ing a holis­tic view of hard­ware and soft­ware as an inte­grat­ed system.

The result is a tech­ni­cal­ly val­i­dat­ed and ful­ly func­tion­al demon­stra­tor that impres­sive­ly demon­strates the poten­tial of intel­li­gent dri­ver assis­tance in micro­mo­bil­i­ty. This show­cas­es our abil­i­ty to trans­late inno­v­a­tive research approach­es into robust, prac­ti­cal, and scal­able sys­tem solu­tions. For the next devel­op­ment step—toward prod­uct matu­ri­ty and mar­ket access—the right strate­gic frame­work and part­ner­ships are now crucial.

The project was fund­ed under the Cen­tral Inno­va­tion Pro­gram for SMEs (ZIM) of the Fed­er­al Min­istry for Eco­nom­ic Affairs and Cli­mate Action. Our devel­op­ment part­ner was the West Sax­on Uni­ver­si­ty of Applied Sci­ences in Zwick­au, Depart­ment of Main­te­nance and Acci­dent Analysis.