Deploying Intelligent Industrial Robots for Smart Manufacturing: Moma Intelligence Gives Robots a "Smart Brain" | Oasis Vitality

Moma Intelligence completed a Pre-A round of tens of millions of RMB in 2022, with Oasis Capital as the sole investor. The company is currently raising a new round.

At the 2023 "Robotics+" conference, Wang Hong, deputy director of the Equipment Industry Division I at China's Ministry of Industry and Information Technology, reported that in 2022, China's industrial robot production reached 443,000 units, with installed volume exceeding 50% of the global total — marking nine consecutive years as the world's largest market.

Despite this massive installed base, the intelligence level of industrial robots still lags. For instance, when switching production lines, industrial robots require redeployment, consuming significant engineering time and effort.

Domestic robot deployment work is dominated by system integrators. Robot deployment demands extensive time and engineering debugging, with rising engineer costs putting considerable cost pressure on integrators. How to improve industrial robot intelligence and deployment service efficiency is an urgent problem facing the industry.

Moma Intelligence focuses on the pain points of industrial robot deployment, providing intelligent solutions that help manufacturing enterprises achieve flexible production while reducing total lifecycle costs and improving efficiency.

Factory environments are complex. Engineers must repeatedly debug industrial robot trajectories based on on-site conditions. This debugging typically consumes hundreds to over a thousand hours; for large production lines, deployment cycles can stretch to 12-24 months.

Moma Intelligence has developed a proprietary cognitive intelligence algorithm training platform. Embedded with AI-based optimal trajectory strategy training algorithms for industrial robots, it can deploy optimal robot trajectories for production environments, achieving unmanned deployment. This allows Moma Intelligence to reduce industrial robotic arm deployment time from dozens or hundreds of hours to just a few hours or even under one hour, helping enterprises achieve intelligent flexible production under personalized market demands.

From a technical perspective, existing deployment methods rely primarily on teach programming, essentially a low-code approach. However, this method lacks precision and stability, still requires engineers on-site for extensive debugging, struggles with high-precision operation scenarios, and cannot adapt to multi-variety, small-batch flexible production needs. Every product change demands engineers redeploy motion trajectories, force control, positioning accuracy, and other parameters — time-consuming and inefficient.

Trajectory planning typically requires experienced robot application engineers to debug first in virtual simulation software such as ABB's RobotStudio and Siemens' Process Simulate. Even after programs transfer to physical robotic arms, substantial on-site re-debugging remains necessary.

From a customer usability perspective, this manual trajectory planning approach is largely an experience-based trial-and-error method. Moma Intelligence instead uses AI algorithms to rapidly find optimal paths, comprehensively considering production cycle time, obstacle avoidance, force control, impact, and other multi-dimensional objectives to directly output optimal production trajectory strategies.

Moma Intelligence explains that manual debugging produces a fixed trajectory curve from point A to point B. If any obstacle or interference constraint suddenly appears along this curve, manual re-debugging is required. AI-generated results, by contrast, represent a motion strategy — robots can autonomously find paths on-site, proactively adapt to production line and product SKU changes, and automatically complete trajectory programming.

Moma Intelligence's proprietary cognitive intelligence algorithm training platform resembles an autonomous driving simulation training platform. Its goal is to train what amounts to an L4 system for robots. Through on-site sensors uploading data in real-time to a TCU (trajectory control unit), millisecond-level autonomous path judgment and decision-making occurs, with commands then issued to the robot controller, giving robotic arms autonomous adaptability.

From the perspective of current market delivery models, industrial robot solution deliveries typically include only 10-15 SKUs at most. Beyond this quantity, delivery prices escalate further. The reason: deploying industrial robots takes longer, involves greater complexity, and incurs higher labor costs.

If integrators or end-user enterprises adopt Moma Intelligence's robot intelligence software, they can potentially reduce overall robot solution delivery costs while improving integrator delivery efficiency — a win-win across the industry chain.

Moma Intelligence can also help enterprises digitally preserve production-related process data and expertise online, continuously optimizing algorithms to improve production efficiency and further reduce dependence on robot debugging engineers.

According to IFR statistics, in 2021 China's manufacturing robot density was 322 units per 10,000 workers, compared to 1,000 in South Korea, 399 in Japan, and 397 in Germany.

Currently, high robot-density countries including the United States and Japan have seen leading companies enter industrial robot intelligence. Examples include Intrinsic, incubated by OpenAI and Google. In December 2022, Intrinsic acquired Open Source Robotics Corporation (OSRC), the commercial entity behind the industrial robot open-source operating system ROS, along with OSRC-SG, an independent company established by OSRC in Singapore.

As domestic industrial robot deployment density increases, deployment challenges will become more pronounced. The industrial value of industrial robot intelligence technology will thus become increasingly prominent.

Moma Intelligence's flexible loading/unloading technology has already been applied in the semiconductor and 3C industries. The company notes that this technology applies not only to multi-variety, small-batch industries such as IC chips and aerospace, but can also rapidly extend to machining, logistics, assembly, and other sectors with growing flexibility requirements.

Current customers include Fortune 500 companies, listed companies, and other leading industrial clients across integrated circuits, automotive manufacturing, aerospace, defense, and consumer goods industries. Specific clients include a global top chip design company, a global top server manufacturer, a listed semiconductor packaging and testing company, and Apple supply chain companies.

The Moma Intelligence team currently numbers over 40 people, with more than 60% holding master's degrees or higher. The team includes a national "Thousand Talents" expert, with members hailing from Tsinghua University, State University of New York, Free University of Berlin, Kyushu University, Shanghai Jiao Tong University, Northwestern Polytechnical University, Harbin Institute of Technology, and other renowned domestic and international institutions. The company maintains AI and robotic arm technology R&D centers in Shanghai and Berlin, with an engineering application center in Ningbo.

Source: 36Kr

Oasis Capital is a new-generation venture capital firm in China, dedicated to discovering the most vital entrepreneurs of the next decade and growing alongside them to create long-term value. "Nurturing Vitality" is Oasis Capital's vision and mission. This vitality represents both the direction of structural transformation in the era and the resilience and evolutionary power of entrepreneurs.

Oasis Capital focuses on early and growth-stage investments, with individual investments ranging from $3 million to $30 million. It concentrates on robotics, artificial intelligence, technology services, and other sectors, supporting China's technology-driven new service upgrade.