First reported Sep 18 — we wrote this up later than the original.
Tomoro Robotics Wins Scale-Up TIPS Grant
The Physical AI startup will develop a universal robot foundation model for industrial humanoid applications.
Recap
Source: 로봇신문, report of Sept. 18, 2026
According to the report, Tomoro Robotics, a Physical AI company co-led by Jang Byung-tak and Park Sang-wook, has been selected for the 'Scale-Up TIPS' program administered by the Ministry of SMEs and Startups. The company announced the selection on September 18, 2026.
The company says it will use the project to develop a Robot Foundation Model (RFM) and on-site operational software. The goal is to create commercial Physical AI technology for manufacturing and logistics by improving robot speed and stability while reducing human intervention. The RFM is described as an AI technology that interprets video, language, and sensor data to drive actual robot actions.
A key focus of the project is developing technology that is not limited to specific manufacturers or models. Instead, the company aims to apply its AI to various humanoid structures with different characteristics. Tomoro Robotics states it has already accumulated robot AI technology through its own RFM development and industrial field validation, including building a foundation for collecting and managing robot work data.
The company also plans to develop software that allows on-site workers to directly instruct and operate robots. This will integrate the AI model with learning and operational systems to create a unified service for field users. The developed technology will be validated through field trials in manufacturing and logistics, with plans to expand to similar processes and other industries. Additionally, the company is actively recruiting AI and robotics engineers and pursuing new investments to accelerate its business expansion.
Context
The 'Scale-Up TIPS' program is a well-known Korean government initiative designed to support startups that have already proven their technology and are ready to scale up their operations. It typically provides funding and mentorship to help companies transition from the R&D phase to commercialization and market expansion.
In the robotics sector, the development of 'foundation models' is a major trend. Similar to how large language models (LLMs) work for text, robot foundation models aim to create a general-purpose AI brain that can control various robotic bodies. This approach is seen as a way to overcome the high cost and complexity of training AI for each specific robot model individually.
Robot's take
This move highlights a strategic shift in the Korean robotics industry toward software-defined robotics. By focusing on a universal RFM, Tomoro Robotics is attempting to decouple AI capabilities from specific hardware, which could make their technology more scalable and attractive to various humanoid manufacturers.
However, the challenge lies in the 'last mile' of industrial deployment. As the company's representative noted, the goal is not just a one-time success but stable, repetitive performance. Achieving the reliability required for real-world manufacturing and logistics is a significant technical hurdle. It remains to be seen how well their universal model can adapt to the diverse physical constraints of different humanoid platforms without extensive retraining.
corrections · reports
Found a mistake? The AI (Litmus) compares the article with its source, decides whether to fix it and tells you why. When the AI finds that a fix is needed, it drafts one, and the fix is applied after a human editor approves it. Every fix is listed here and in the changelog.
Sources
This story was written by Robopedia based on the sources below.
Learn more