Traffic light optimization with low penetration rate vehicle trajectory data.

Traffic light optimization is known to be a cost-effective method for reducing congestion and energy consumption in urban areas without changing physical road infrastructure. However, due to the high installation and maintenance costs of vehicle detectors, most intersections are controlled by fixed-time traffic signals that are not regularly optimized. To alleviate traffic congestion at intersections, we present a large-scale traffic signal re-timing system that uses a small percentage of vehicle trajectories as the only input without reliance on any detectors. We develop the probabilistic time-space diagram, which establishes the connection between a stochastic point-queue model and vehicle trajectories under the proposed Newellian coordinates. This model enables us to reconstruct the recurrent spatial-temporal traffic state by aggregating sufficient historical data. Optimization algorithms are then developed to update traffic signal parameters for intersections with optimality gaps. A real-world citywide test of the system was conducted in Birmingham, Michigan, and demonstrated that it decreased the delay and number of stops at signalized intersections by up to 20% and 30%, respectively. This system provides a scalable, sustainable, and efficient solution to traffic light optimization and can potentially be applied to every fixed-time signalized intersection in the world.

Learning naturalistic driving environment with statistical realism.

For simulation to be an effective tool for the development and testing of autonomous vehicles, the simulator must be able to produce realistic safety-critical scenarios with distribution-level accuracy. However, due to the high dimensionality of real-world driving environments and the rarity of long-tail safety-critical events, how to achieve statistical realism in simulation is a long-standing problem. In this paper, we develop NeuralNDE, a deep learning-based framework to learn multi-agent interaction behavior from vehicle trajectory data, and propose a conflict critic model and a safety mapping network to refine the generation process of safety-critical events, following real-world occurring frequencies and patterns. The results show that NeuralNDE can achieve both accurate safety-critical driving statistics (e.g., crash rate/type/severity and near-miss statistics, etc.) and normal driving statistics (e.g., vehicle speed/distance/yielding behavior distributions, etc.), as demonstrated in the simulation of urban driving environments. To the best of our knowledge, this is the first time that a simulation model can reproduce the real-world driving environment with statistical realism, particularly for safety-critical situations.

Dense reinforcement learning for safety validation of autonomous vehicles.

One critical bottleneck that impedes the development and deployment of autonomous vehicles is the prohibitively high economic and time costs required to validate their safety in a naturalistic driving environment, owing to the rarity of safety-critical events1. Here we report the development of an intelligent testing environment, where artificial-intelligence-based background agents are trained to validate the safety performances of autonomous vehicles in an accelerated mode, without loss of unbiasedness. From naturalistic driving data, the background agents learn what adversarial manoeuvre to execute through a dense deep-reinforcement-learning (D2RL) approach, in which Markov decision processes are edited by removing non-safety-critical states and reconnecting critical ones so that the information in the training data is densified. D2RL enables neural networks to learn from densified information with safety-critical events and achieves tasks that are intractable for traditional deep-reinforcement-learning approaches. We demonstrate the effectiveness of our approach by testing a highly automated vehicle in both highway and urban test tracks with an augmented-reality environment, combining simulated background vehicles with physical road infrastructure and a real autonomous test vehicle. Our results show that the D2RL-trained agents can accelerate the evaluation process by multiple orders of magnitude (103 to 105 times faster). In addition, D2RL will enable accelerated testing and training with other safety-critical autonomous systems.

Intelligent driving intelligence test for autonomous vehicles with naturalistic and adversarial environment.

Driving intelligence tests are critical to the development and deployment of autonomous vehicles. The prevailing approach tests autonomous vehicles in life-like simulations of the naturalistic driving environment. However, due to the high dimensionality of the environment and the rareness of safety-critical events, hundreds of millions of miles would be required to demonstrate the safety performance of autonomous vehicles, which is severely inefficient. We discover that sparse but adversarial adjustments to the naturalistic driving environment, resulting in the naturalistic and adversarial driving environment, can significantly reduce the required test miles without loss of evaluation unbiasedness. By training the background vehicles to learn when to execute what adversarial maneuver, the proposed environment becomes an intelligent environment for driving intelligence testing. We demonstrate the effectiveness of the proposed environment in a highway-driving simulation. Comparing with the naturalistic driving environment, the proposed environment can accelerate the evaluation process by multiple orders of magnitude.

Safety assessment of highly automated driving systems in test tracks: A new framework.

Among the three major safety assessment methods (i.e., simulation, test track, and on-road test) for highly automated driving systems (ADS), test tracks provide high fidelity and a safe and controllable testing environment. However, due to the lack of realistic background traffic, scenarios that can be tested in test tracks are usually static and limited. To address this limitation, a new safety assessment framework is proposed in this paper, which integrates an augmented reality (AR) testing platform and a testing scenario library generation (TSLG) method. The AR testing platform generates simulated background traffic in test tracks, which interact with subject ADS under test, to create a realistic traffic environment. The TSLG method can systematically generate a set of critical scenarios under each operational design domain (ODD) and the critical scenarios generated from the TSLG method can be imported into the AR testing platform. The proposed framework has been implemented in the Mcity test track at the University of Michigan with a Level 4 ADS. Field test results show that the proposed framework can accurately and efficiently evaluate the safety performance of highly ADS in a cost-effective fashion. In the cut-in case study, the proposed framework is estimated to accelerate the assessment process by 9.87×104 times comparing to the on-road test approach.