Paper #1
Formalizing traffic rules for uncontrolled intersections
Abstract – One of the challenges in designing autonomous vehicles (AV’s) is driving around humans (i.e. drivers, cyclists, pedestrians, etc.) In particular, the AV’s and the humans must have a common set of traffic rules to follow. Changing the driving rules to accommodate AV’s seems impractical, given the high cost of human education, road infrastructure, and lack of standards for AV technologies. An alternative is to adapt AV’s to current traffic rules. Considering the ambiguity and complexity of natural language, the problem is to develop a formal version of traffic rules that is machine-understandable. The challenge is to preserve the human understanding of the rules as much as possible in the model. In this paper, we present a new approach to formalize and implement traffic rules. We use California’s DMV driver handbook as a working example. Our approach provides a straightforward mapping from the rules in the handbook to its formal model, and from the model to its implementation. To demonstrate the efficiency of the implementation, we develop a traffic controller for CARLA that computes the right-of-way in real time and sends the correct action to the vehicles. CARLA is an open-source simulator for autonomous driving research. The autopilot vehicles in CARLA are simulator-controlled agents that are used to create a traffic environment. Our controller computes the right-of-way of all (autopilot, AI-controlled or manually controlled) vehicles at an uncontrolled intersection. Then it sends the correct action to autopilot vehicles. Our controller makes the behaviour of autopilot vehicles more realistic, compared to CARLA’s default FIFO controller, by modeling more right-of-way rules. This also improves the throughput of the traffic through the intersection. We implement the traffic rules in the logic programming paradigm of Answer Set Programming (ASP).
Video presentation