Astribot Launches SmoothRL: A New Era in Online Reinforcement Learning for Robots
- Published
- Source
- leaderobot.com
Astribot has introduced SmoothRL, an online reinforcement learning framework validated through real robot tasks. This technology addresses a critical challenge in embodied intelligence: how robots can learn and improve while executing tasks in real-time. Traditional models struggle with precision in real-world applications, often failing in details like positioning and force control despite extensive training.
The significance of SmoothRL lies in its ability to allow robots to receive feedback through real interactions, adjusting their strategies based on success or failure. Unlike conventional offline training, which pauses for data collection and model updates, SmoothRL enables continuous operation, thus enhancing the robot's learning process without disrupting its tasks. This shift marks a pivotal change in how robots are trained, moving from reliance on pre-collected data to ongoing adaptation based on real-world experiences.
Initial tests with the Astribot S1 robot show promising results, with task success rates significantly improving across various activities. However, challenges remain for widespread adoption, including high data collection costs and the need for reliable reward mechanisms. The future of robotic learning may evolve into a continuous optimization process, moving away from fixed capabilities post-deployment.
Original report
From Execution to Evolution: Online Reinforcement Learning Transforming Embodied Robot Development
leaderobot.com · LeaderobotThis briefing is an independently written summary based on publicly available reporting and is provided for industry information and news discovery. The original report and source publication are credited and linked where applicable. RobotToday does not claim ownership of third-party source material.
Rights concerns? Contact [email protected] with the relevant URL and details. We will review the matter and take appropriate action where warranted.
More related news
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
- 11
- 12
Related Suppliers
Astribot (Stardust Intelligence) 星尘智能
China
Shenzhen, China humanoid-robot maker founded late 2022; its dual-arm S1 platform reaches over 10 meters per second with 0.1 mm repeatability.
Algomatic Dynamics
Japan
Tokyo physical-AI startup founded 2026 building AI multi-fingered robot hands and bipedal robots; launched with a JPY 5 billion round from DMM.com.
BodenAI 宁波博登智能科技有限公司
China
Ningbo training-data firm founded 2019 that teleoperates humanoids and arms in its Physical AI Dojo and sells annotated robot datasets and data tooling.
ThoughtForge AI
United States
AI-powered robotic control platform using Active Inference for 99%+ accuracy in dynamic industrial environments.
Temple Allen Industries
United States
US manufacturer of AI-driven robotic surface preparation systems for aerospace and defense, eliminating manual sanding ergonomic risks.
Chengdu Humanoid Robot Innovation Center Co., Ltd. 成都人形机器人创新中心有限公司
China
Chengdu-registered humanoid robot R&D company behind the Gongga-1 humanoid, the open-source RW-02OP wheeled-biped platform and Raydiculous-1 planning software
Fatfish Technology Pte Ltd
Singapore
Singapore's FatFish AI builds Visual Action Intelligence software, led by its CAMP One engine, for cross-brand robot perception and action.
TGW Logistics GmbH
Austria
Austrian foundation-owned warehouse automation systems integrator with 4,600+ employees and €1.07B revenue, serving global fulfillment centers.
Continuous Composites
United States
Coeur d'Alene, Idaho firm founded 2012 that builds CF3D robotic cells printing continuous-fiber composite parts; closed a $17M Series A in 2021.
GMO AI & Robotics Corporation
Japan
Tokyo-area GMO AI & Robotics, founded June 2024 as a GMO Internet Group unit, is Japan's authorized Unitree Robotics distributor for humanoid robots.
Telefactor Robotics
United States
US specialist in teleoperated robot vision, dexterous grippers, and control interfaces for bomb disposal and hazardous environment robots.
Learning Robots
France
Learning Robots is a Paris-Saclay CNRS spin-off making AlphAI, an educational robot and software for teaching artificial intelligence.
Robotics needs a service framework.
RSF defines a common language for robot service capability, lifecycle operations, certification pathways, and service-provider networks.
Join the RobotToday community on LinkedIn
Daily robotics news, in-depth analysis, conference highlights, and discussions with professionals worldwide.
