April 18, 2015      

“Nuclear companies are reportedly using robotics to deliver safer, faster, and more cost-effective solutions for nuclear decommissioning around the world, that is forecast to be worth $370 billion between now and 2030.” —Financial Times

Big problem

A human’s bio-system is killed off within an hour of entering into close proximity of the containment vessel in the No. 1 reactor at Fukushima.

OK, humans are goners, but what about a robot’s electronic system?

snake robot

Research has demonstrated that no robot is robust enough or can be hardened enough — without encasing it in concrete — to withstand the same No. 1 reactor for very long.

Although the world seems busy at building humanoid or mobile or tractor-driven or snake-like or flying disaster robots capable of dealing with post-nuclear-event environments like Fukushima, the fact is that none can survive for very long when in the presence of gamma rays, or worse, neutrons, that abound in hostile, nuclear kill zones like the No. 1 reactor.

What happens to a robot’s innards when it’s confronted by what is called a Total Ionizing Dose (TID), which is basically a robot taking a bath in a sea of nuclear radiation. Think, human getting a really bad sunburn, and then some!

A robot’s essential system components like microprocessors, microcontrollers, actuators, and transistors do not fair too well in TID, which in turn cause operations to fail, like the inability of an instruction set to execute properly.

The question that pops instantly to mind: can any robot be ruggedized or in some way hardened to survive the nuclear rigors of Fukushima?

Challenge and opportunity

Knowing what stops a robot in its tracks is what Arthur Witulski and his engineering mates at Vanderbilt University do all day. Their latest research paper seems right on the money when considering the recent total failure of the snake-like robotic probe sent in to survey the No. 1 reactor (April 10, 2015).

Witulski et al research, funded by the U.S. Defense Threat Reduction Agency (DTRA), is titled: A Comprehensive Program for Investigation of Radiation Effects in Robots Used in Mitigation of Nuclear Disasters.

The robot sent into the reactor by TEPCO (Tokyo Electric Power Co.) lasted three hours, having completed less than a third of its scheduled ten-hour mission, before prematurely stopping inside the reactor.

TEPCO gave up on recovering the robot, cut the robot’s tether and left it there. Without a post-mortem on the robot, we may never know why it failed.

A second, identical robot was sent in on April 16th, and it completed the assigned task. How much longer it could have lasted in the TID bath within the reactor may well be found out upon examining the robot.

However, according to the Japan Times: “After its exploratory trip, which will make the robot extremely radioactive, technicians plan to store it in a shielded box. They have no plans to reuse it.”

If the Japanese authorities are looking to robots to provide heavy assistance in cleaning up the Fukushima disaster area over the next forty years of decommissioning, then this robot failure most certainly presents a massive concern.

Since there are thousands of nuclear reactors worldwide, any one of which could experience the same fate as Japan’s tragedy, the world undoubtedly shares the same concerns about robot assistance during nuclear incidents.

See related: Back to Fukushima with a Snake Robot

Challenge in search of a solution

At a press showing back in February, there were high hopes for the 2-foot-long (60-centimeter) robot developed by Hitachi’s nuclear partner, Hitachi-GE Nuclear Energy.

According to the Japan Times: previous computer simulations indicated that “all of the fuel rods in No. 1 probably melted and pooled at the bottom of the containment chamber.” The Hitachi-GE robot would have been the first to confirm those simulations.

In February, Hitachi-GE engineer Yoshitomo Takahashi had hoped that “Depending on how much data we can collect from this area, I believe (the robot) will give us a clearer vision for future decommissioning.”

Witulski feels certain that TEPCO and its partner were well aware of the potential for gamma rays to disrupt and then to stall the robot. It was a risk that they had to take. So far, neither TEPCO nor Hitachi have made any public statements about radiation-induced degradation to any of the robot’s electronics or even why they think that the robot failed.

Total-ionizing-dose robustness on robot sensors

Measurements of the ionizing radiation in the basement of No. 1 reactor have hit astounding highs of over 5,000,000 microSieverts per hour! By comparison, the average U.S. adult receives 3,600 microSieverts in one year from all sources (well below government-established safety limits); a single dose of 2,000,000 microSieverts means severe radiation poisoning (sometimes fatal).


A robot’s electronics, in the same No. 1 basement environment, are bombarded per hour with enough radiation to kill a human three times over.

Most of that radiation is in the form of gamma rays (similar to X-rays) in that they can pass through most any material. It takes lead or several feet of concrete to stop them.

How nimble would a robot be if its effectiveness was hampered by lead or concrete?

Witulski and his fellow Vanderbilt engineers have researched the effect of gamma rays on a robot’s sensors; sensors being critical for a robot to perform range sensing through dust, smoke, water vapor, etc., and to have a sense of itself in relation to the environment around it.

In short, if a robot is incapable of good to excellent sensing capabilities in the basement of No. 1 reactor, it’s helpless to do any work there.

In their research paper, Range-Finding Sensor Degradation in Gamma Radiation Environments, they illustrate sensor radiation degradation not only in operational failure, but also in changes in the sensor transfer function for range-finding sensors (infrared, sonar using time of flight, and laser rangefinder using triangulation).

In a telephone interview, Witulski described the gamma rays as high-energy particles depositing their energy into the oxides of integrated circuits, which, in turn, degrade and then stop a robot’s sensing functions.

It seems apparent that in any post-radiation-disaster environment, a robot that is needed to perform work might find it impossible for its sensory systems to survive the conditions there.