MR-X: Mantis launches fenceless dual arm robot for shared factory floors

Mantis Robotics has introduced the MR-X, a biomimetic dual arm system designed to operate alongside human workers without a traditional safety cage. The company positions it as a first in class machine that blends speed with safety, building on the MR-1 which is certified to ISO 10218 and ISO 13849 and has already demonstrated that industrial speeds can be attained without fencing. The MR-X expands that proven approach into a new form factor, signaling a bold push toward broader use of fenceless robotics on real factory floors. Gerry Vannuffelen, CEO of Mantis, frames the MR-X as more than a novelty. He says this is evidence that the company’s safety architecture can support fast and productive collaboration in environments that mix humans and robots. “Do not mistake this for another humanoid robot. It is a first in class dual arm robot that outperforms both cobots and humanoids in terms of speed and safety,” he said.

At the heart of MR-X is the patented Mantis SafetyCore platform, a reflex system designed to give the robot full, continuous awareness of its surroundings. The claim is not simply that the machine can move quickly; it is designed to maintain safety as it ramps up to industrial speeds without the need for traditional fences. Mantis emphasizes that MR-X builds on the MR-1’s track record of fenceless operation, backed by major automation players including Amazon, and then extends the technology to a more capable form factor. The company describes MR-X as capable of fixed installations as well as mobile manipulator deployments, a design choice intended to offer flexibility across a range of applications from assembly cells to material handling.

For plant leaders, the appeal is straightforward: a dual arm platform that can operate in a shared workspace with humans, potentially increasing throughput where tasks require both dexterity and strength. Yet the deployment reality is nuanced. The MR-X promises industrial speeds, but the exact cycle times and throughput figures remain undisclosed in the rollout. What is clear is that integration will matter as much as speed. The MR-X requires seamless alignment with existing line controls, safety interlocks, and sensor systems to maintain its fenceless posture without compromising worker safety. In practice, that means the line must be instrumented with robust perception, calibration, and real time collision avoidance logic, plus a workflow that assigns human operators to tasks that benefit from judgment and adaptability rather than repetitive precision alone.

From a practitioner’s lens, several constraints and tradeoffs stand out. First, the promise of fenceless operation depends on a safety architecture that can reliably detect humans and adapt behavior in real time; that often translates into careful cell design, sensor placement, and ongoing maintenance to prevent occlusions or sensor drift. Second, the integration burden is nontrivial: PLCs, vision systems, and safety interlocks must be harmonized with the MR-X’s reflex based control loop. Third, the technology is positioned as augmenting rather than replacing craft labor. It can shoulder repetitive, high speed tasks or awkward handling, while skilled trades such as linemen, inspectors, welders, and other craft roles apply expertise where judgment, quality checks, and complex assembly are required. Finally, the cited intent to mass deploy hinges on demonstrated real world performance beyond controlled demonstrations, with deployment data gradually translating into measurable improvements in cycle time and overall throughput.

The MR-X launch sends a clear strategic signal: if a dual arm platform can run at industrial speeds without fences, then the ROI equation for many manufacturing lines pivots on integration discipline and operator training as much as on robot capability. The company’s emphasis on SafetyCore, plus MR-1’s ISO certifications and Amazon’s backing, sets a pathway for broader adoption, but the next test will be on actual production floors where line changes, maintenance windows, and human factors determine whether speed translates into sustained throughput gains. As Vannuffelen puts it, this is not just a robot launch; it is a stance that mass deployment of fenceless robotics is a practical reality, not a far off promise.