Karsten Kreutz and Julian Eggert, "Robust Car Following Models require Explicit Reaction Times", Intelligent Transportation Systems Conference (ITSC) 2022, 2022.

Car following models like the Intelligent Driver Model (IDM) describe the longitudinal behavior of an ego-car depending on a leading vehicle. Especially the IDM has been applied within a broad range of mobility-related tasks like the analysis of traffic phenomena, microscopic traffic simulations as well as for single vehicle behavior prediction. Although car following models can be formulated with explicit delays (e.g. in form of system reaction times), the IDM is generally used in form of a time-continuous differential equation without delay. One reason is that in early IDM parameter calibration analysis work with real trajectories, it was concluded that explicit delays do not improve the overall model trajectory reconstruction accuracy. While this is true for slowly changing or quasistationary traffic conditions, it is questionable for trajectories with abruptly changing behavior like sudden braking. Especially in risky conditions this type of behavior prevails, making it relevant to capture car behavior as accurately as possible. For this purpose, in this paper, based on an extended set of real trajectories, we analyze the question of the advantage of explicit delays or reaction times. We show that indeed both reconstruction and prediction errors benefit from considering explicit delays in the IDM, improving and recommending its usage for safety-relevant scenarios like required for behavior prediction in intelligent ADAS.