Robots are super useful in many ways, especially when it comes to doing things that we humans don’t want to do. Whether it’s dirty, dangerous or just downright boring, over the centuries we’ve invented automatons to lighten the load.
There are lots of places that humans want to explore, but are inhospitable to visit in person. Robots make great tools for allowing us to investigate extreme environments like the deep ocean, volcanic plumes, or even other planets in our solar system.
One of those places is nuclear storage bunkers, where radiation means that humans can only spend a limited amount of time inside. Humans wear dosimeters (which measure radiation dose) and if the dose approaches a certain level, they must pack up their things and leave. Radiation doses are kept deliberately low to ensure the safety of workers, but it’s inconvenient to stop working halfway through a job.
My work in the Robotics Group at the University of Manchester focuses on the total characterisation for remote observation in nuclear environments, or TORONE for short. Put simply, I put lasers on robots to find out what objects are made of in places humans can’t go.
Robots are far more resistant to radiation, can work in the same environment for longer and help to reduce the radiation dose to human workers. Hurrah!
When exploring new places, we want to know what stuff is made of. For humans, simply looking at an object will tell us what it is because of the colour and texture, even if we’re looking at a photograph. We can do this because of all the examples of materials we see in our everyday lives.
But how can a robot know what something is made of? And in the environments we work in, how can a human know what something is made of when seeing it for the first time? In this instance, we turn to scientific instruments that can tell us about the elements and how they are put together.
For elemental composition, laser-induced breakdown spectroscopy (LIBS) is used. A high intensity laser pulse creates a plasma (an ionised gas that gives off light—think the flame of a candle but without burning) on the surface. The colour of the plasma is dependent on the elements of the target material. Using a spectrometer, it is possible to uniquely identify each element.
LIBS is used on the Mars rover Curiosity, where it identifies the composition of rocks and soils on the Martian surface. It has also been used in forensics to detect stolen art and forgeries, by analysing the types of paints used without damaging the possibly priceless painting.
So far, so awesome. But sometimes knowing the elemental composition is not enough and you need more detail. For example, rust is a compound of iron and oxygen but seeing iron and oxygen when using LIBS does not necessarily tell you it is rust. So what to use then? Step in Raman spectroscopy.
Raman spectroscopy can identify chemical bonds, and in that case, you detect the particular bond between iron and oxygen that identifies the material as rust. Raman spectroscopy uses a low power laser (it can be achieved using a simple laser pointer if you are skilful enough) and a sensitive spectrometer, as the signal is comparatively weaker than LIBS.
Raman spectroscopy is a very powerful tool and used in many industries, such as the pharmaceutical industry, where it can test the composition of unknown medicines in a non-destructive way.
For both instruments we use off-the-shelf research grade lasers which have been modified to make them as small and as lightweight as possible. The robots deployed may be small (no bigger than 50 cm x 50 cm x 50 cm) but they’re hard as nails research robots, known as Clearpath Jackals. They’re made of stainless steel with chunky 4-wheel drive action, making them rugged and weatherproof. Vroom!
Much like humans before the advent of Google Maps, robots find it difficult to navigate new environments and they need to explore new surroundings to figure out where they can go and how to get there. When we send these robots into extreme environments, we’re really interested in what’s there and so the robots are normally controlled manually. Once an operator is happy they have seen everything, the robot can be made to autonomously fill in any missing parts of the map.
With a complete map, the robot can be made to take samples of materials using its on-board instruments, usually in locations where a human can’t tell what something is made of. All this useful data is stored in a database, where it can be analysed further. The maps and data generated can be used to plan future missions, and if humans wish to enter an environment, they can practice using VR equipment.
The work we are doing at TORONE will combine different strands of robotic technology to form a seamless jigsaw—a ‘total characterisation’ picture. The future use of nuclear energy in the UK and internationally is dependent on the ability to characterise highly radioactive environments for both efficient decontamination and decommissioning, as well as in the design of new nuclear fission reactors as well as fusion reactors.