The COVID-19 pandemic made robot manufacturers explore the idea of combining mobile robotics with UV-C
light to automate the disinfection processes. But performing this process in an optimum way introduces some
challenges: on the one hand, it is necessary to guarantee that all surfaces receive the radiation level to ensure
the disinfection; at the same time, it is necessary to minimize the radiation dose to avoid the damage of the
environment. In this work, both challenges are addressed with the design of a complete coverage path planning
(CCPP) algorithm. To do it, a novel architecture that combines the glasius bio-inspired neural network (GBNN),
a motion strategy, an UV-C estimator, a speed controller, and a pure pursuit controller have been designed.
One of the main issues in CCPP is the deadlocks. In this application they may cause a loss of the operation, lack
of regularity and high peaks in the radiation dose map, and in the worst case, they can make the robot to get
stuck and not complete the disinfection process. To tackle this problem, in this work we propose a preventive
deadlock processing algorithm (PDPA) and an escape route generator algorithm (ERGA). Simulation results
show how the application of PDPA and the ERGA allow to complete complex maps in an efficient way where
the application of GBNN is not enough. Indeed, a 58% more of covered surface is observed. Furthermore, two
different motion strategies have been compared: boustrophedon and spiral motion, to check its influence on
the performance of the robot navigation.
