HiWonder is behind some of the coolest robots we have covered. The WonderDuck is a new robot that can walk, recover from a fall and do a whole lot more. It has an open hardware design and is ready for reinforcement learning. This robot has two legs and walks like a duck. It can use its beak to pick up items.
HiWonder plans to make the operating code and training projects open. Details are sketchy but we will find out soon enough what this robot has to offer.
A tiny robotic duck is taking its first steps into the world of AI!
Meet WonderDuck from HiWonder, a walking robot designed to bring robotics research closer to real-world experiments.
With open hardware, code and training projects, developers can explore reinforcement learning, simulation and deploying trained models on physical robots.
Available in four colours, it’s a playful way to learn serious robotics.
#WonderDuck #HiWonder #Robotics #EmbodiedAI #ReinforcementLearning
#AI@AlbertoEMachado @baski_LA @efipm @sallyeaves @gvalan @chboursin @helene_wpli @tewoz @aure79lien @Guillaume_Rio @jeffkagan @odisseiaalfa @andresvilarino @erikbryn @YuHelenYu @wil_bielert @marek_rosa @michaelqtodd @BetaMoroney @MaiaGabunia @AnthonyRochand @bimedotcom
— Amitav Bhattacharjee (@bamitav) November 9, 2026
[HT]