In the past, we have covered robotics project that involve teaching robots how to automatically adapt to injuries like animals. As Jean-Baptiste Mouret explains, the Intelligent Trial and Error algorithm enables robots to adapt to damage in am matter of minutes. The team has extended the algorithm to allow robots to learn to compensate for damages while completing their tasks.
More like this ➡️ here
Towards semi-episodic learning for robot damage recovery
This means robots of the future will be able to perform autonomously for a longer period of time and adapt to damage. The above video shows a 6-legged robot performing locomotion tasks without a left middle leg.
[Source]
*Our articles may contain aff links. As an Amazon Associate we earn from qualifying purchases. Please read our disclaimer on how we fund this site.
