Legged robots are edging closer to day-to-day use in warehouses and industrial sites, forcing regulators to confront risks that traditional automation never posed. A new report from Interact Analysis shows safety standards pivoting toward fall behavior, application-level risk, and cyber resilience as these machines leave the lab and enter live operations.
Why Legged Robots Create New Safety Problems
According to the report, the mechanics of legged robots introduce hazards that do not exist with statically stable systems such as fixed automation or wheeled vehicles. Clara Sipes, a market analyst at Interact Analysis and author of the report, notes that dynamically stable machines, including humanoids, depend on continuous power and active control to remain upright. When power is lost, these robots do not simply stop; they fall.
That distinction matters operationally. Falls introduce residual risk to workers, infrastructure, and goods, particularly in shared spaces where robots and people operate side by side. Unlike forklifts or AGVs, whose behavior during shutdown is predictable, legged robots can collapse unpredictably depending on posture, load, and surface conditions.
To address these issues, new standards such as ISO 26058-1 and ISO 25785-1 have been developed to explicitly cover both statically and dynamically stable mobile robots. In parallel, technical report TR R15.108 examines safety considerations specific to bipedal, quadrupedal, and self-balancing wheeled robots, reflecting growing recognition that “mobile robotics” is no longer a single risk category.
From ‘Collaborative Modes’ to Context-Based Risk
One of the more consequential shifts highlighted in the report is the removal of blanket references to “collaborative modes.” Earlier frameworks tended to classify robots themselves as collaborative or non-collaborative. Updated guidance instead evaluates risk at the application level, recognizing that the same robot can present very different hazards depending on task, environment, speed, and human proximity.
Under the revised approach, safety assessments are expected to account for the full operating context—facility layout, traffic patterns, interaction points, and failure scenarios—rather than relying on a robot’s design label. This change aligns with how many logistics operations already think about risk: as a property of workflows and systems, not individual machines.
The report also points to expanding requirements around cyber resilience. As robots become more network-connected and software-defined, cybersecurity failures increasingly translate into physical safety risks. European regulatory developments, including the Cyber Resilience Act, the AI Act, and the updated Machinery Regulation, are converging toward a unified framework that links functional safety, cybersecurity, and AI governance.
Where Early Compliance Quietly Creates Operating Leverage
As safety, cybersecurity, and AI governance converge into a single regulatory surface, early alignment is becoming an operational choice rather than a legal one. Organizations that treat emerging standards as input into system architecture, rather than as documentation to be layered on later, are better positioned to integrate legged robots into mixed human environments without constraining throughput or flexibility. Recent regulatory guidance makes clear that safety validation will increasingly be tied to real operating conditions, software updates, and network exposure, not static certifications. Designing for that reality now reduces friction later, particularly as robots shift from contained pilots into continuously evolving, software-driven assets on the floor.
