Automated, modular, asynchronous and locally controlled material flow systems promise high routing flexibility in production lines because their conveying modules can be reconfigured without reprogramming PLCs. However, if such material flow systems comprise cycles and different routes, they may exhibit undesirable deterministic chaotic inter-arrival times, which can lead to conveying bottlenecks when approaching maximum capacity. Since existing analytical models have not been practically adopted for planning material flow systems, an approach for detecting deterministic chaotic inter-arrival times during production is proposed. It employs the Hough transform to identify trajectories in inter-arrival time phase space. The approach is tested with a laboratory double belt conveyor system, in which non-deterministic behavior is minimized. Results are compared with a previously published analytical model. It is shown that the proposed approach is able to detect deterministic chaotic inter-arrival times for the test cases. Phase trajectories are only partly identified. Future research should test and compare different line detection algorithms for their influence on the approach’s robustness in practical production environments.