Arm brings together more than 80 partners to introduce a common robotics framework, aiming to simplify integration and speed up automation in physical industries. Early supporters include AWS and Siemens.
Robotics development has often been a patchwork of incompatible hardware and software, making integration slow and costly. Arm is trying to change that with its new Arm Total Design for Physical AI, a project aimed at setting shared standards for automated systems in physical industries. More than 80 companies have already joined, according to an official announcement, signaling broad interest across the sector.
Industry analysts estimate that by the 2030s, physical industries such as mining, agriculture, manufacturing, and transport will require around $200 billion in annual compute capacity.
Arm’s partner list includes AWS, Hugging Face, Siemens, NXP, QNX, Unitree Robotics, and Liquid AI, covering hardware, sensors, AI models, runtime software, and compute silicon. Their shared goal is to tap into the $200 billion annual compute market projected for physical industries like mining, agriculture, manufacturing, and transport by the 2030s.
For developers and manufacturers, the lack of standards has meant high integration risk and slow scaling from prototype to deployment. Arm’s framework is designed to remove these hurdles, making it easier to move from proof-of-concept to real-world use. The company’s chief architect, Richard Grisenthwaite, has published a technical manifesto outlining the rationale and inviting others in the sector to help shape the standard.
Industry collaboration
The Robotics Capability Framework is positioned as a "common language" for defining and comparing robotic systems, linking real-world use cases to behavioral requirements, outputs, and system constraints.
In the automotive sector, Arm has already shown how this approach can work. By collaborating with AWS, Google, HERE, RemotiveLabs, and Siemens, Arm enabled software teams to develop and test digital cockpit solutions on the Arm Zena CSS platform before the hardware was available. Using digital twins and virtual platforms, teams can speed up development and reduce risk for both autonomous transport and robotics projects.
Arm is now asking for technical input from the wider engineering community. As physical AI systems become more advanced, the company argues that shared standards and open collaboration will become even more important.
Operational impact
For content creators, publishers, and digital media operators, more standardized robotics and AI systems could mean smoother integration with publishing workflows, automated content production, and data collection. For industries investing in AI-driven automation, having a shared way to benchmark and compare systems could speed up adoption and lower risk.
The framework’s long-term success will depend on continued support from both established companies and startups. The initial group of partners is broad, but the real challenge will be putting these standards into practice across different industries and environments.
By opening the framework to community input, Arm is positioning itself as a central player in setting standards for physical AI. If the company can keep momentum and deliver on interoperability, it could play a key role in the next phase of industrial automation. Whether this coalition can hold together as companies move from theory to deployment remains to be seen.
Arm, founded in 1990 and based in Cambridge, UK, reported $2.7 billion in revenue in its most recent fiscal year. Its technology powers more than 250 billion chips worldwide. The company’s move into physical AI and robotics marks a shift beyond its traditional focus on mobile and cloud, targeting sectors where standardized automation could unlock significant economic value in the coming decade.
