RealAIGym: Education and Research Platform for Studying Athletic Intelligence
Abstract:
Traditional robots today (such as the ones used in factories) have a fixed base and are fully actuated under their operating conditions.
However, modern robots inspired by animals are not bound to one place and are always under-actuated. Like animals, these robots can perform dynamic movements, demonstrate compliance, and are robust to contact during their movements. The interest in dynamic robot behaviors has increased significantly due to the impressive athletic behaviors shown by robots developed by Boston Dynamics, MIT Mini Cheetah, Agility Robotics etc. This gives rise to the need for canonical robotic hardware setups for studying underactuation and comparing learning and control algorithms for their performance and robustness. These hardware setups and the accompanying software should be affordable, open and accessible.
Similar to OpenAIGym and Stable Baselines which provide simulated benchmarking environments and baselines for Reinforcement Learning (RL) algorithms, the concept of RealAIGym is introduced with a set of reproducible robotic hardware platforms, for establishing a baseline for the application of dynamic control algorithms on real hardware.
Website: https://dfki-ric-underactuated-lab.github.io/real-ai-gym/
project:
underactuated-lab: https://robotik.dfki-bremen.de/en/research/research-facilities-labs/underactuated-lab
Видео RealAIGym: Education and Research Platform for Studying Athletic Intelligence канала German Research Center for Artificial Intelligence
Traditional robots today (such as the ones used in factories) have a fixed base and are fully actuated under their operating conditions.
However, modern robots inspired by animals are not bound to one place and are always under-actuated. Like animals, these robots can perform dynamic movements, demonstrate compliance, and are robust to contact during their movements. The interest in dynamic robot behaviors has increased significantly due to the impressive athletic behaviors shown by robots developed by Boston Dynamics, MIT Mini Cheetah, Agility Robotics etc. This gives rise to the need for canonical robotic hardware setups for studying underactuation and comparing learning and control algorithms for their performance and robustness. These hardware setups and the accompanying software should be affordable, open and accessible.
Similar to OpenAIGym and Stable Baselines which provide simulated benchmarking environments and baselines for Reinforcement Learning (RL) algorithms, the concept of RealAIGym is introduced with a set of reproducible robotic hardware platforms, for establishing a baseline for the application of dynamic control algorithms on real hardware.
Website: https://dfki-ric-underactuated-lab.github.io/real-ai-gym/
project:
underactuated-lab: https://robotik.dfki-bremen.de/en/research/research-facilities-labs/underactuated-lab
Видео RealAIGym: Education and Research Platform for Studying Athletic Intelligence канала German Research Center for Artificial Intelligence
Показать
Комментарии отсутствуют
Информация о видео
5 июля 2022 г. 17:46:50
00:06:05
Другие видео канала
BaSynaos - Online-Prozessplanung für Intralogistik und Produktion im Basissystem für Industrie 4.0Kollaborierende Roboter und klassische Industrierobotik? | Dr. Tim Schwartz | DFKIWhy do we need an explainable AI?DFKI auf der HANNOVER MESSE 2023FASTER: Explore space faster and safer (short version)CoMem – AI to assist daily work in companiesCoMem – KI zur Assistenz der täglichen Arbeit im UnternehmenTransTerrA: Coyote III crater trialsANT - navigation system enables walking systems to explore rough inclined terrainBIONIC – Intelligent sensor networks reduce physical stress30 Jahre DFKI – Open House – Trailer KurzfassungWas ist Deep Learning?DFKI auf der CeBIT 2016ANT – Novel Navigation System for Multi-legged Robots, second phaseProject WALL-ET – Assistance in supermarkets and warehouses through social, cognitive robotsWie denkt ein Informatiker? – DFKI beim Digitaltag 20213D Dual-Reality Benutzerschnittstelle / Guided ABIntelligent space robotics: scientist Wiebke Brinkmann speaks on the EU-funded research in BremenSELFIE: Self-Verification of Electronic Systems30 Jahre DFKI – Open House – Trailer Langfassung