Home Robot Safety Is All About Relationships
Original reporting by IEEE Spectrum (Robotics)
As advanced personal care robots prepare to enter homes, assisting with everything from elder care to daily chores, the question of their safety has become paramount. The International Standards Organization (ISO) is finalizing a long-anticipated update to its 12-year-old safety standard, ISO 13482, a revision deemed critical as domestic humanoid robot makers pivot from lab prototypes to consumer products. This update aims to define acceptable robot behavior in the complex, uncontrolled environments of real families.
The governance gap
Yet, according to technology policy researcher Jae-Seong Lee, while the proposed standard addresses hazard identification and risk assessment, it critically falls short. The core engineering challenge, Lee argues, isn't just collision avoidance but the "bidirectional" nature of human-robot interaction. Safety, in this context, isn't a fixed property of the machine alone; it emerges dynamically from the relationship. While the technical community understands these relational hazards, the ISO standard, in its current form, fails to convert this knowledge into binding compliance criteria, test methods, or enforceable rules for domestic autonomy. It acknowledges the complexity but pushes unresolved issues into advisory language, delaying critical decisions about what constitutes safe relational behavior.
Lee emphasizes a profound "governance gap": who decides whose movements define normal, whose risk thresholds are acceptable, and whose definition of safe judgment gets embedded in these foundational requirements? Without systematic representation from the people most affected, such as older adults who are primary intended users, there's a risk that safety assumptions are "baked in" before the market, regulators, and users have a chance to question them, making future revisions far more difficult.
The revised ISO 13482 standard for personal care robots marks an important step in acknowledging the intricate challenges of human-robot collaboration in domestic settings. Yet, as researcher Jae-Seong Lee highlights, it falls short of providing the binding compliance criteria and enforcement mechanisms necessary to truly address "relational safety"—the dynamic, bidirectional interplay where human and robot mutually influence each other's behavior and perceived safety. The current update, while progressive in its recognition of these complexities, risks deferring crucial decisions, potentially embedding safety assumptions into products before a robust, inclusive framework can be established for their widespread deployment.
Defining Future Safety
The implications of this governance gap extend far beyond immediate physical safety. Without explicit, enforceable rules for domestic autonomy, critical value judgments about "normal" behavior, acceptable risk thresholds, and safe judgment are implicitly made by a limited group, often without the systematic input of those most affected, such as children or older adults. This oversight risks baking in biased or incomplete safety baselines that will be profoundly difficult to revise once domestic robots become widespread in homes globally. The future of truly safe and beneficial personal care robotics depends on a paradigm shift: moving beyond machine-centric measurements to system-level relational assurance, where the human is not merely an external factor but an integral component in defining the safety envelope. This necessitates a more inclusive and explicit standard-setting process to ensure that these evolving technologies genuinely serve humanity in all its diversity.