Radiation-Hardened Wi-Fi Chip Could Unleash Wireless Robots in Nuclear Reactors

Last updated: 2025

Robots have been cleaning up Fukushima for over a decade — tethered by cables that tangle, snag, and limit where they can go. A new Wi-Fi receiver chip developed at the Institute of Science Tokyo survived 500 kilograys of radiation, potentially cutting the cord for nuclear decommissioning robots at a moment when 200 reactors globally will reach end-of-life within 20 years.

Why nuclear robots are still wired in 2025

Post-Fukushima robots operate almost entirely over physical LAN cables — and those cables are a genuine operational liability inside a damaged reactor. The problem is not engineers failing to imagine wireless; it's that radiation destroys conventional electronics fast enough to make wireless unreliable.

After the 2011 Fukushima Daiichi disaster, robotic systems became essential for mapping and characterising the site's interior. The challenge was immediate: standard Wi-Fi hardware simply doesn't survive long enough to be useful inside an active or recently shut-down reactor. LAN cables became the fallback — functional, but operationally cumbersome. They tangle around debris, limit a robot's range of motion, and create failure points in environments where human intervention to untangle a cable is not an option.

Yasuto Narukiyo, a graduate student working with advisor Atsushi Shirane and KEK researcher Masaya Miyahara, presented a potential solution at the IEEE International Solid-State Circuits Conference (ISSCC) in San Francisco in February: a 2.4 GHz Wi-Fi receiver redesigned from the ground up to survive inside a nuclear reactor.

The timing matters. According to a 2024 study published in ScienceDirect, of 204 reactors that have already been closed worldwide, only 11 plants above 100 megawatts capacity have been fully decommissioned. The other 193 remain in various stages of shutdown — and 200 more reactors will hit end-of-life within the next two decades. That is an enormous, underserved market for specialised robotics.

How the radiation-hardened chip actually works

The receiver achieves radiation hardness through three deliberate design changes: replacing vulnerable PMOS transistors with inductors, enlarging transistor gate geometry, and aggressively reducing total transistor count. Together, these modifications shifted the chip's radiation tolerance from "consumer electronics" to "inside an operating reactor."

The root vulnerability in standard chips is the oxide layer inside silicon MOSFETs (metal-oxide semiconductor field-effect transistors — the fundamental switching elements in nearly all modern electronics). When gamma rays hit this oxide layer, they knock loose positive charges that become trapped, progressively degrading the transistor's switching behaviour and introducing signal errors.

The team attacked this problem at the component selection level. PMOS transistors — where current flows via positive charge carriers — are doubly vulnerable: trapped charges accumulate in both the oxide layer and at the interface between the oxide and the semiconductor substrate. The cumulative effect pushes the transistor toward an "off" state, degrading circuit function. The redesigned receiver minimises PMOS use entirely, substituting passive inductors (coils that store energy in magnetic fields rather than relying on oxide-sensitive switching) wherever possible.

NMOS transistors, where electrons carry the current, proved more resilient. Positive charges trapped in the oxide are partially cancelled by negative charges accumulating at the interface — a natural compensation mechanism. The team leaned into this, keeping NMOS elements where active switching was unavoidable.

The third change was geometric. The transistor's gate — the control electrode that switches current on and off — becomes more susceptible to radiation damage as it shrinks. The counterintuitive fix: make the gates physically larger. Longer, wider gates distribute radiation-induced charge over more material, reducing its localised impact on device performance.

How tough is 500 kGy — and is it enough?

500 kilograys is roughly 1,000 times the radiation tolerance required for space-rated electronics, and over 3,000 times the dose that causes a standard KUKA robotic arm to fail. The chip maintained functional Wi-Fi reception after this exposure, with signal gain dropping by only 1.5 decibels — a small penalty for surviving conditions that would destroy almost any commercial hardware.

To put the numbers in perspective:

The chip clears the minimum bar for nuclear deployment. Performance before irradiation was comparable to standard commercial Wi-Fi receivers; after 300 kGy and then 500 kGy exposure, the gain loss of 1.5 dB is operationally acceptable. Narukiyo himself characterises the receiver as "hardened enough" — the emphasis now shifts to improving range, data throughput, and companion transmitter hardware.

What still needs to be solved before wireless nuclear robots are real

A Wi-Fi receiver alone enables one-way signal reception. Nuclear robot control requires two-way communication — and the transmitter half of that equation remains unsolved. An earlier prototype transmitter was destroyed at 300 kGy, well below the 500 kGy threshold the receiver survived.

Transmitters face a harder physics problem than receivers. Generating a Wi-Fi signal requires producing high levels of current, which demands more complex circuitry and more transistors — each one a radiation vulnerability. The team's current receiver works in part because they stripped component count down aggressively. A transmitter cannot make the same tradeoffs without losing the ability to generate a usable signal.

The research group is investigating alternative semiconductor materials — including diamond electronics — to achieve the hardness level needed. Diamond has exceptional radiation tolerance due to its wide bandgap (the energy gap that determines how electrons move through a material), but fabricating diamond-based circuits at useful complexity remains a significant manufacturing challenge.

Until a transmitter matches the receiver's durability, the system cannot replace wired control links. Robots operating in these environments need reliable command channels, not just telemetry uplinks.

What This Means for Robotics in Hazardous Environments

Wireless connectivity is one of the most significant remaining constraints on what nuclear and extreme-environment robots can actually do. Removing the cable tether would open up access to confined spaces, allow faster robot deployment, and reduce the risk of mechanical failure from snagged lines.

For the robotics industry, the decommissioning backlog represents a concrete, fundable market — one where operators have clear budget motivation (human radiation exposure liability) and where conventional solutions are demonstrably insufficient. The 200 reactors reaching end-of-life in the next 20 years are not a distant scenario; decommissioning timelines typically stretch 20-40 years from shutdown, meaning procurement decisions for robotic systems are being made now.

The chip-level breakthrough here is also relevant beyond nuclear. Any environment with intense electromagnetic or ionising radiation — particle accelerators, certain aerospace applications, mining operations near radioactive ore bodies — faces similar wireless connectivity constraints. A proven radiation-hardened Wi-Fi architecture creates a foundation that other industries can build on.

For teams already exploring used industrial robots for hazardous environment deployment, the communications layer is often the binding constraint. Hardware that can physically survive extreme conditions is increasingly available; the wireless link is the missing piece.

This research is early-stage — a single chip presented at a conference, not a shipping product. But the underlying approach is methodologically sound, the performance numbers are credible, and the application demand is unambiguous. The next milestone to watch: a transmitter that survives 500 kGy. When that exists, wireless nuclear robotics moves from research to engineering.

Frequently Asked Questions

What radiation level can the new Wi-Fi receiver withstand?

The receiver developed at the Institute of Science Tokyo survived a total radiation dose of 500 kilograys (kGy) — the minimum threshold required for electronics operating inside a nuclear reactor during decommissioning. Signal gain degraded by only 1.5 decibels after full exposure, which is considered operationally acceptable for robot control applications.

Why can't nuclear decommissioning robots use standard wireless communications?

Standard Wi-Fi electronics fail at radiation doses well below what nuclear reactors produce. A KUKA industrial robotic arm, for example, was documented failing at approximately 164.55 grays (0.165 kGy). Space-rated electronics — among the most radiation-tolerant commercially available — are typically rated for 100 to 300 grays over three years, roughly 1,000 times less than nuclear reactor requirements.

How does radiation damage electronics, and how did the team prevent it?

Gamma radiation traps positive charges in the oxide layers of MOSFET transistors, degrading switching performance and causing errors. The team mitigated this by replacing vulnerable PMOS transistors with inductors, keeping NMOS transistors where active switching was required (they are more naturally radiation-resilient), and enlarging transistor gate dimensions to distribute radiation-induced charge effects over more material.

Is wireless nuclear robot control ready to deploy?

Not yet. The receiver is functional and hardened, but a matching transmitter — needed for two-way robot control — has not yet achieved the same radiation tolerance. An earlier transmitter prototype was destroyed at 300 kGy. The team is exploring diamond semiconductors and other wide-bandgap materials to solve the transmitter problem before a complete wireless system can be fielded.

How large is the nuclear decommissioning robotics market?

According to a 2024 study, 200 reactors globally will reach end-of-life within the next 20 years, joining the 204 already closed — of which only 11 large plants have been fully decommissioned. The decommissioning process typically spans decades and carries significant radiation exposure liability, creating sustained demand for robotic systems that can operate in these environments without human presence.

If the transmitter problem gets solved, which reactor decommissioning project do you think deploys wireless robots first — Fukushima or Europe's growing shutdown backlog?

The radiation-hardened Wi-Fi receiver chip demonstrates that wireless nuclear robot communications is an engineering problem, not a physics impossibility. With 200 reactors queued for decommissioning and the transmitter challenge now clearly defined, the next few years of research will determine whether robots at Fukushima and beyond finally cut the cord.