Abstract
Microbial auditing has showed a significant promise in assessing the levels of cleanliness for surfaces beyond visual inspection in different public environments. Additionally, current microbiological methods for cleanliness auditing are highly laborious and lab-oriented, making measurements for large areas unpractical and unfeasible. In this thesis, we explore the development of an automated biosensor that can be attached to autonomous mobile robots as a payload to quantify the degree of microbial contamination density. The degree of microbial contamination density is calculated by the biosensor using Relative Light Units (RLU) at discrete sampled points. The RLU values of the biosensor are validated with benchmark standards from a commercial system. For cleaning personnel, the framework identifies general areas of lower cleanliness levels that require intensive cleaning by assigning a Pass-Fail factor and cleanliness score. For cleaning robots, three interpolation techniques, Nearest Neighbour (NN), Radial Basis Function (RBF), and Ordinary Kriging (OK), are proposed to estimate and map the global distribution of microbial contamination density of an environment. The accuracy of the estimation techniques are compared through the proposed Leave Out One (LOO) and Biosensor Validation (BV) methods, using Root Mean Square Error (RMSE) as a metric for comparison. The framework generates specific cleaning waypoints based on areas of lower cleanliness levels. Four optimisation algorithms, Simulated Annealing (SA), Genetic Algorithm (GA), Ant Colony Optimisation (ACO), and Particle Swarm Optimisation (PSO) are used to generate the shortest route for the robot to cover all waypoints. The performance of the algorithms are then compared. The proposed framework’s usefulness is evaluated through real-world experimental trials at different locations. The obtained results showed that the proposed framework effectively measures and maps the global distribution of microbial contamination density while also charting an optimal path for cleaning robots during a cleaning process.