Real-Time Avoidance of Ionising Radiation Using Layered Costmaps for Mobile Robots

Abstract

Humans in hazardous environments take actions to reduce unnecessary risk, including limiting exposure to radioactive materials where ionising radiation can be a threat to human health. Robots can adopt the same approach of risk avoidance to minimise exposure to radiation, therefore limiting damage to electronics and materials. Reducing a robot’s exposure to radiation results in longer operational lifetime and better return on investment for nuclear sector stakeholders. This work achieves radiation avoidance through the use of layered costmaps, to inform path planning algorithms of this additional risk. Interpolation of radiation observations into the configuration space of the robot is accomplished using an inverse distance weighting approach. This technique was successfully demonstrated using an unmanned ground vehicle running the Robot Operating System equipped with compatible gamma radiation sensors, both in simulation and in real-world mock inspection missions, where the vehicle was exposed to radioactive materials in Lancaster University’s Neutron Laboratory. The addition of radiation avoidance functionality was shown to reduce total accumulated dose to background levels in real-world deployment and up to a factor of 10 in simulation.

CARMA II: A ground vehicle for autonomous surveying of alpha, beta and gamma radiation

“…This is because, due to its high penetration, gamma radiation can be easily detected and does not carry the same contamination risks as alpha and beta. Moreover, the method used has already been proven to be effective in a previous publication by the authors (West et al, 2022).…”

“…Other research works have focused on constructing a map of gamma radiation within an environment while minimising the radiation dose received by the robot (Groves et al, 2021;West et al, 2022). Including radiation information within the navigation system's costmaps to give the robot the ability to avoid regions with high radiation was first developed by Groves et al (2021) and was later proven to reduce the dose that the robot received by West et al (2022). This approach works by introducing additional path planning cost to traverse areas with elevated gamma radiation dose rate.…”

“…to the costmap produced by move-base-flex using the radeye and radiation-layer packages (West et al, 2022). Parameters related to standard and custom packages can be configured using the dynamicreconfigure 3 package.…”

“…Different parameters such as radiation thresholds and inflation radius can be tuned by the user for better performance in radiation informed navigation. The concept of the layered costmap with radiation layers is detailed in our previous work (West et al, 2022) and is not a contribution of this paper. In the case of CARMA II, the modified radiation-layer package takes into account alpha, beta and gamma radiation and integrates these measurements into the relevant costmaps.…”

“…As was the case in Experiment 2, in Figure 9D the peak dose rate associated with the robot in close proximity to the gamma source is decreased for the avoiding case, with total accumulated dose being decreased as a result. As additional path planning cost is linearly proportional to dose rate, it is important to note that only when the dose rate is above a certain threshold, based on numerous factors, does the robot take any considerable action West et al (2022). This ensures that coverage is prioritised, with only the significant threat to the health of the robot leading to a safer route.…”