ABSTRACT
IoT (Internet-of-Things)-powered devices can be exploited to connect vehicles to smart city infrastructure, allowing vehicles to share their intentions while retrieving contextual information about diverse aspects of urban viability. In this paper, we place ourselves in a transient scenario in which next-generation vehicles that are able to communicate with the surrounding infrastructure coexist with traditional vehicles with limited or absent IoT capabilities. We focus on intersection management, in particular on reusing existing traffic lights empowered by a new management system. We propose an auction-based system in which traffic lights are able to exchange contextual information with vehicles and other nearby traffic lights with the aim of reducing average waiting times at intersections and consequently overall trip times. We use bid propagation to improve standard vehicle trip times while allowing emergency vehicles to free up the way ahead without needing ad hoc system for such vehicle, only an increase in their budget. The proposed system is then tested against two baselines: the classical Fixed Time Control system currently adopted for traffic lights, and an auction strategy that does not exploit traffic light coordination. We performed a large set of experiments using the well known MATSim transport simulator on both a synthetic Manhattan map and on a map we built of an urban area located in Modena, Northern Italy. Our results show that the proposed approach performs better than the classical fixed time control system and the auction strategy that does not exploit coordination among traffic lights.
ABSTRACT
With the advent of IoT, cities will soon be populated by autonomous vehicles and managed by intelligent systems capable of actively interacting with city infrastructures and vehicles. In this work, we propose a model based on reinforcement learning that teaches to autonomous connected vehicles how to save resources while navigating in such an environment. In particular, we focus on budget savings in the context of auction-based intersection management systems. We trained several models with Deep Q-learning by varying traffic conditions to find the most performance-effective variant in terms of the trade-off between saved currency and trip times. Afterward, we compared the performance of our model with previously proposed and random strategies, even under adverse traffic conditions. Our model appears to be robust and manages to save a considerable amount of currency without significantly increasing the waiting time in traffic. For example, the learner bidder saves at least 20% of its budget with heavy traffic conditions and up to 74% in lighter traffic with respect to a standard bidder, and around three times the saving of a random bidder. The results and discussion suggest practical adoption of the proposal in a foreseen future real-life scenario.