The cleanup of a decommissioned nuclear reactor is a slow, mechanical choreography performed in environments hostile to both human tissue and silicon. Since the 2011 disaster at Fukushima Daiichi, robots have been the primary agents of characterization and decontamination inside damaged reactor buildings. Yet these machines remain largely tethered by local area network (LAN) cables — physical lifelines that frequently tangle in collapsed structures, snag on debris, and stall operations in environments where every delay carries radiological and financial cost. A team at the Institute of Science Tokyo has now demonstrated a wireless receiver designed to survive inside the reactor itself, a step toward cutting the cord for good.

Presented at the IEEE International Solid-State Circuits Conference in San Francisco, the receiver endured a total ionizing dose of 500 kilograys — a threshold orders of magnitude higher than the standards applied to space-grade electronics, which are typically rated for far lower cumulative exposure. The work, led by graduate student Yasuto Narukiyo alongside colleagues Atsushi Shirane and Masaya Miyahara, targets a deceptively simple objective: replace the physical tether with a robust Wi-Fi control link so that a snagged cable does not stall a critical mission in a high-stakes environment.

Why the tether persists

The reason nuclear-cleanup robots still trail cables has less to do with locomotion or sensing than with communication. Radiation degrades semiconductor devices through well-understood mechanisms — ionizing particles create charge traps in oxide layers, shift transistor thresholds, and eventually render circuits inoperable. Standard commercial Wi-Fi chipsets fail at doses far below what the interior of a damaged reactor accumulates over a decommissioning campaign. Space-hardened electronics, while more resilient, are designed for the comparatively moderate radiation environment of low Earth orbit or interplanetary missions, not for sustained proximity to spent fuel or melted corium. The result is that operators default to wired connections, accepting the operational penalty of tangled cables rather than risking a communication blackout mid-task.

The 500-kilogray figure reported by the Tokyo team is significant because it approaches the cumulative doses expected in the most contaminated zones of facilities like Fukushima Daiichi, where robots must eventually retrieve fuel debris. Achieving that tolerance in a Wi-Fi receiver — a mixed-signal circuit that must maintain precise frequency selectivity and signal integrity — represents a distinct engineering challenge compared with hardening simpler digital logic. The specific circuit-design techniques the team employed to reach this threshold were presented at the conference, signaling that the approach has passed at least the peer-review bar of one of the field's most selective venues.

A growing queue of reactors

The implications extend well beyond disaster recovery. Nuclear power plants are finite structures; they eventually reach the end of their operational lives and require meticulous dismantling before a site can be repurposed. Of the 204 reactors closed worldwide, only a small fraction have been fully decommissioned. The remainder sit in various stages of defueling, decontamination, and structural demolition — processes that can stretch across decades. As the global fleet ages, the queue of facilities awaiting retirement will lengthen, and with it the demand for robots capable of navigating radioactive labyrinths without the limitation of a physical leash.

Historically, decommissioning timelines have been shaped as much by the availability of suitable tooling as by regulatory or financial constraints. Remote-handling technology matured through programs at sites such as Sellafield in the United Kingdom and Three Mile Island in the United States, but each generation of robots has confronted the same bottleneck: reliable communication in high-radiation fields. Hardening the communication layer, rather than the manipulator or the sensor suite, addresses what has long been the weakest link in the chain.

Whether the Tokyo receiver can transition from a conference demonstration to a deployable module inside an actual reactor building remains an open question. Integration challenges — antenna design for cluttered metallic environments, network reliability under fluctuating dose rates, qualification testing to nuclear-industry standards — sit between the laboratory result and operational use. At the same time, the pressure to accelerate decommissioning is unlikely to ease. The tension between a maturing but still laboratory-stage technology and an expanding global inventory of aging reactors will shape how quickly untethered robots move from proof of concept to standard practice.

With reporting from IEEE Spectrum Robotics.

Source · IEEE Spectrum Robotics