The most expensive part of any factory robot isn't the arm, the gripper, or the software. It's the cage. The safety fencing, light curtains, and perimeter scanners that keep humans out of the robot's work cell can double the total cost of deployment. Devol Robots, a startup with one foot in Malaysia and another in San Francisco, is betting it can build a robot that doesn't need one.
Its proposition is a next-generation robotic arm built around force-control actuators and what it calls embodied AI. The idea is to make a machine that can sense its environment through touch and adjust its movements in real time, allowing it to work safely in the same space as a human operator [Devol Robots, Unknown]. It's a hardware-first answer to a problem the industry has mostly tried to solve with cameras and software.
A hardware wedge into collaborative robotics
Devol's stated mission is to design and manufacture force-control robots that adapt to their environment and seamlessly interact with humans [Devol Robots, Unknown]. While many competitors in the collaborative robotics, or 'cobot', space retrofit standard industrial arms with sensors and software, Devol is starting from the actuator. The company claims its proprietary force-control technology gives its robots a human-like sense of touch, allowing for delicate manipulation and, crucially, an inherent ability to stop or yield upon unexpected contact.
This is a classic hardware wedge. By owning the fundamental mechanical layer, Devol aims to achieve a level of safety and dexterity that software wrappers on rigid arms struggle to match. The company is incorporated in Malaysia as Devol Robots Sdn. Bhd., described as a homegrown robotic arm company, while its CEO, Sze Yuan Cheong, operates from San Francisco [LinkedIn, Unknown]. This dual structure suggests a strategy of leveraging lower-cost manufacturing and engineering talent in Southeast Asia while maintaining commercial and technical leadership in a major robotics hub.
The team and its trans-Pacific footprint
The public team picture is still developing, but the pieces point to a technical founding group with US educational ties building in Asia. CEO Sze Yuan Cheong is an Iowa State University graduate [LinkedIn, Unknown]. The company's Malaysian entity lists engineers with backgrounds from the University of Michigan, and the company has also incorporated a separate entity in Singapore, signaling a focus on the regional manufacturing corridor [LinkedIn, Unknown] [sgpbusiness.com, Unknown].
| Name | Role | Location | Background |
|---|---|---|---|
| Sze Yuan Cheong | Co-Founder & CEO | San Francisco, USA | Iowa State University [LinkedIn, Unknown] |
| Tingshiuan Wu | Team Member | Morgan Hill, USA | University of Michigan [LinkedIn, Unknown] |
| Ooi Kai Ping | Robotics Engineer | Malaysia | Not specified [LinkedIn, Unknown] |
The company closed an undisclosed pre-seed round in May 2024 [Crunchbase, May 2024]. The lack of public detail on funding size or lead investors is typical for a hardware startup at this stage, where capital intensity makes stealth a strategic choice.
Where the bet gets hard
Building a new robotic arm from scratch is one of the hardest ventures in hardware. The incumbents are entrenched, supply chains are complex, and sales cycles into manufacturing are long. Devol's bet rests on a few critical assumptions that will determine its fate.
- Performance vs. Price. Its force-control system must be significantly better than the sensor-and-software kits available for Universal Robots or Fanuc arms, at a price point that doesn't erase the savings from removing safety cages.
- The AI promise. The 'embodied AI' claim needs to materialize as real-world adaptability, not just marketing. A robot that learns from physical interaction is a profound software challenge on top of the hardware one.
- Commercial pathfinding. The company must choose its first beachhead application wisely. Precision assembly? Packaging? Machine tending? Each has different tolerance and payback requirements.
The competitive landscape is dense with well-funded players pursuing similar goals of human-robot collaboration through various technical routes. Devol's differentiation is its insistence that a new kind of muscle, not just a better brain, is the key.
For a sense of the scale, consider a typical collaborative robot work cell today. A mid-range cobot arm might cost $35,000. The safety fencing, controllers, and installation can add another $25,000. If Devol's arm costs $45,000 but requires no cage, the customer saves $15,000 upfront and gains valuable floor space. That's the unit economics argument in a nutshell. The company isn't just selling a robot; it's selling deleted square footage.
To succeed, Devol must eventually outmaneuver Universal Robots, the Danish pioneer that created the cobot category and now dominates it with a vast ecosystem of integrators and end-of-arm tools. UR achieved its position by making robots simple to deploy. Devol's challenge is to prove that its more complex hardware ultimately makes deployment even simpler, and safer, for the applications that matter most.
Sources
- [Devol Robots, Unknown] Our Mission | Devol Robots | https://www.devolrobots.com/aboutus
- [LinkedIn, Unknown] Sze Yuan Cheong - Co-Founder & CEO @ Devol Robots | https://www.linkedin.com/in/sze-yuan-cheong-b981a567/
- [Crunchbase, May 2024] Devol Robots - Crunchbase Company Profile & Funding | https://www.crunchbase.com/organization/devol-robots
- [LinkedIn, Unknown] Tingshiuan Wu - Devol Robots Sdn. Bhd. | https://www.linkedin.com/in/wutingshiuan/
- [sgpbusiness.com, Unknown] DEVOL ROBOTS PTE. LTD. (202418834K) - Singapore Company | https://www.sgpbusiness.com/company/Devol-Robots-Pte-Ltd
- [LinkedIn, Unknown] Ooi Kai Ping - Robotics Engineer - Devol Robots Sdn. Bhd. | https://www.linkedin.com/in/ooi-kai-ping-59199b250/