Dexforce Uses Virtual Data to Cut Humanoid Robot Training Costs
Dexforce is using synthetic scenes to train humanoid robots, betting that simulation can reduce reliance on costly human demonstrations.
Daniel Krauss is a technology writer and editor at RobotsBeat, covering the fast-moving intersection of robotics, artificial intelligence, and industrial automation. His work focuses on how intelligent machines are reshaping manufacturing, logistics, and enterprise operations, with particular attention to humanoid robots, physical AI, and large-scale industrial systems. With a clear, analytical style, Daniel translates complex engineering and AI concepts into precise, readable reporting for a global audience. He closely follows major robotics companies, automation platforms, and research labs, tracking how software intelligence is moving out of the cloud and into physical systems on factory floors and in real-world environments. At RobotsBeat, Daniel Krauss contributes in-depth guides, news analysis, and explainers designed for professionals who want signal over hype. His writing emphasizes real deployments, economic impact, and the strategic decisions shaping the future of robotics and AI-driven industry.
Dexforce is using synthetic scenes to train humanoid robots, betting that simulation can reduce reliance on costly human demonstrations.
Boston Dynamics shows how machining, assembly and repair support Atlas development, with a goal of cutting robot build time from a day to an hour.
Amazon’s Physical AI Toolchain combines AWS infrastructure and NVIDIA robotics tools to support training, simulation and real-world deployment.
Counterpoint forecasts humanoid robot DRAM demand will grow more than 20x by 2030, as average memory per robot rises from 19GB to 39GB.
UBTECH and Xiamen C&D plan a joint venture targeting at least 10,000 humanoid robot sales by 2030 across logistics, property and public services.
Tesla is scaling Optimus production ahead of a late-2027 public-sales target, but reliability and useful autonomy will determine the rollout.
XPENG introduces its YOYO robotaxi brand and opens registration in China, with overseas trials and external IRON robot sales planned for 2027.
Stuttgart’s SCALAR and COBRAS projects will develop coordinated construction robots for timber assembly and reusable structures, backed by about €8 million.
A palm-sized robot from UCLA and Michigan uses snapping elastic limbs to hop across six surfaces, climb small steps and swim with added fins.
Ministry of Awe’s Awebot brings a remote-controlled artwork to Philadelphia’s streets, using character design and movement to spark public interaction.