In the real world, most of biological systems that follow Markov process have internal states which are unobservable, so called hidden states. However, although the states could not be directly observed, events emitted under any hidden states would often be observable. It infers that a lot of biological systems might be simulated by hidden Markov model (HMM). To date, many studies tried to apply the HMM for monitoring and describing of animal behaviors. In this study, we attempted to build a HMM that emulates the traveling behavior of subterranean termites in the artificial tunnels with varying width and curvature, based on the empirical data obtained from our previous study that explored the relationship between subterranean termite's movement efficiency and the geometric pattern of their tunnels.