Fuzzy-Bayesian Expert System for Assistance in Bike Mechanical Issues

Cycling is a popular recreational activity and mode of transportation that offers numerous health benefits and environmental advantages. Individuals of all skill levels can benefit from cycling, from beginners to seasoned professionals. However, as with any sport or activity, cyclists face inherent risks and challenges, ranging from safety concerns to performance optimization. Maintenance and mechanical fixing are standard activities in cyclists but could become so complex if it is the first time the user will perform them. Moreover, they could also be dangerous for professional or amateur users. However, it could be easier to fix mechanical issues in bikes if the user has the tools information, the repair instructions, and the probability of fixing the problem so that the user better let an expert fix it. Alternatively, Expert systems have exhibited skills assisting in different science fields, allowing users to query knowledge with forward and backward chaining to obtain information and answers to diagnose a problem and how to solve it. Additionally, Bayesian theory and fuzzy logic allow working with conditional probabilities and imprecise knowledge in expert systems. In this research, we propose a Fuzzy-Bayesian expert system that helps amateur and professional cyclists diagnose and decide whether to fix the issue themselves or search for an expert. Our proposal, developed in Python, uses UPAFuzzySystems to describe fuzzy rules and Twilio to allow SMS communication to send the report to the user, enabling him to maintain the information at hand while repairing the bike.