The Mall as Laboratory: Demystifying the Robot
In the summer of 2025, the RAI Institute converted a section of the CambridgeSide mall in Cambridge, Massachusetts, into a public-facing robotics exhibition. The initiative, led by executive director Marc Raibert, placed advanced autonomous machines in the corridors of a working shopping center — near food courts, retail storefronts, and the ordinary foot traffic of a weekend afternoon. The premise was straightforward: if robots are to operate alongside people, the public needs to encounter them outside the curated settings of trade shows and viral videos.
The pop-up featured a museum tracing the history of robotics alongside live demonstrations of systems such as the ANYmal, a quadrupedal robot developed by Swiss firm ANYbotics, and mobile platforms built by the RAI Institute itself. Visitors could observe these machines navigating real-world obstacles — uneven flooring, curious children, shopping bags left in their path — rather than watching them perform rehearsed routines on a conference stage.
Closing the Perception Gap
The project addresses a persistent asymmetry in how the public understands robotics. Media coverage tends to oscillate between two poles: existential alarm about machines replacing human workers, and breathless enthusiasm about technological omnipotence. Neither frame reflects the current state of the field, where most robots remain limited by the physics of manipulation, the brittleness of perception algorithms, and the sheer unpredictability of unstructured environments. A quadruped that can traverse a warehouse floor may still struggle with a wet tile or a toddler darting across its path.
This gap between narrative and reality is not merely a communications problem. It has practical consequences. Public anxiety can slow regulatory approval for robotic deployments in healthcare, logistics, and municipal services. Conversely, inflated expectations can lead to premature adoption and inevitable backlash when systems underperform. The RAI Institute's mall experiment is, in effect, an attempt to calibrate expectations — to let people form impressions based on direct observation rather than algorithmic feeds.
The approach has precedent. The automotive industry spent decades staging public demonstrations to normalize the idea of self-driving vehicles, from DARPA Grand Challenge events to Waymo's early rider programs in Phoenix. The robotics sector, by contrast, has largely confined its public engagement to controlled demos at CES or academic conferences. Moving into a shopping mall represents a deliberate shift toward the kind of ambient, low-stakes exposure that builds familiarity over time.
A Sociological Experiment in Disguise
Beyond public outreach, the CambridgeSide installation functions as a data-gathering exercise in human-robot interaction, a subfield that has grown steadily as autonomous systems move closer to consumer-facing roles. By observing how shoppers of varying ages and backgrounds react to legged robots — whether they approach, avoid, photograph, or attempt to communicate with them — researchers collect behavioral signals that laboratory settings cannot replicate. The social dynamics of a mall, where attention is divided and context is informal, offer a richer dataset than a university hallway ever could.
This line of inquiry matters because the bottleneck for widespread robotic deployment is increasingly social rather than mechanical. Industrial robots have operated in factories for decades, but those environments are structured, access-controlled, and populated by trained workers. The next frontier — delivery robots on sidewalks, assistive machines in elder care facilities, autonomous platforms in retail — requires coexistence with people who have no training and no obligation to cooperate.
The RAI Institute's experiment raises a question that the robotics industry has been slow to confront directly: whether public acceptance is something that can be engineered through better design, or whether it requires a longer, messier process of cultural negotiation. Hardware improvements and software refinements will continue, but the CambridgeSide pop-up suggests that the more consequential variable may be proximity — the simple, repeated experience of sharing space with a machine and discovering that the encounter is neither threatening nor magical, but ordinary.
With reporting from IEEE Spectrum Robotics.
Source · IEEE Spectrum Robotics
