Exclaim Robotics Closes €4.29M to Build AI Data Center Robots

The big picture: Exclaim Robotics, a Zürich-based startup building robots to maintain AI data centres, has exited stealth mode with a €4.29 million ($4.95 million) pre-Seed round co-led by Founderful and Playfair. The company is building autonomous mobile robots that perform routine repair and maintenance work inside AI data centres.

Why it matters:

• Hazardous maintenance: Next-generation AI data centers will distribute power at 800 volts DC, making manual maintenance tasks like replacing drives or optical transceivers dangerous due to arc-flash risks.
• Staffing shortage: The rapid expansion of AI data center capacity is outpacing the availability of trained technicians, creating a critical gap in physical maintenance capabilities.
• Automation gap: While most data center operations are managed remotely, physical hardware failures still require human intervention, highlighting a need for automated on-site solutions.

How it works:

• Autonomous navigation: Exclaim Robotics’ robots are designed to navigate data halls independently, position themselves at racks, and perform tasks from floor to top rail without requiring modifications to existing infrastructure.
• Initial tasks: The company is focusing on high-volume break-fix tasks, including the replacement and cleaning of optics, SFPs, copper network cabling, and drive replacement.
• Simulation-first training: Robots are trained extensively in simulation to practice tasks more times than possible on physical hardware, then refined on real racks to ensure reliability before deployment.

The catch: Data center environments are inherently complex and cluttered, posing significant challenges for robotic perception and manipulation. The primary difficulty lies in a robot’s ability to accurately understand its surroundings, especially dense cabling, before safely interacting with components. When confidence in a task is low, the system defaults to a remote operator, indicating that full autonomy for all complex scenarios remains a developmental hurdle.