Study of cognitive route search technique for self-driving vehicles

Abstract

In this paper, we introduce Time-Constrained Heuristic Search (TCS) as a route search technique that incorporates a cognitive aspect in self-driving vehicles. TCS is a search method using which we can set the time limit to generate a route in advance. If we consider a situation in which self-driving vehicles will be mass-produced and widely distributed, a large number of vehicles driving on similar routes generated through similar route search methods may cause new types of traffic congestion. This will result in a situation when the optimal route will no longer remain optimal. This is regarded as the social dilemma in the field of game theory and social psychology. Individual rational choice does not correspond with rational choice for a society as a whole. TCS can overcome this by generating a route within a time limit. A longer time limit yields a route more similar to the optimal route. TCS allows us to control the optimality using the time limit available instead of always adopting the optimal route. The most important feature of TCS is that the node expansion in the latter portion of the journey near the destination tends to be minimized to ensure that the search finishes within the prescribed time limit. TCS conducts a search during the initial portion of the journey that is related to the route the vehicle will move on in the immediate future, but does not take into account the latter portion of the route to a great extent. So, TCS can be regarded as the search technique that selects an option to generate immediate benefit. This characteristic is well known as the present-oriented bias or the prospect theory in cognitive psychology. Hence, we can infer that TCS provides an ability of cognitive judgment to self-driving vehicles.