A Data-Informed Analysis of Scalable Supervision for Safety in Autonomous Vehicle Fleets

TL;DR

The present article proposes DISCES, a framework for Data-Informed Safety-Critical Event Simulation, to investigate the practicality of this concept from a dynamic network loading standpoint and characterize the dynamic supervision requirements and thereby scalability of the teleoperation approach.

Abstract

Autonomous driving is a highly anticipated approach toward eliminating roadway fatalities. At the same time, the bar for safety is both high and costly to verify. This work considers the role of remotely-located human operators supervising a fleet of autonomous vehicles (AVs) for safety. Such a 'scalable supervision' concept was previously proposed to bridge the gap between still-maturing autonomy technology and the pressure to begin commercial offerings of autonomous driving. The present article proposes DISCES, a framework for Data-Informed Safety-Critical Event Simulation, to investigate the practicality of this concept from a dynamic network loading standpoint. With a focus on the safety-critical context of AVs merging into mixed-autonomy traffic, vehicular arrival processes at 1,097 highway merge points are modeled using microscopic traffic reconstruction with historical data from interstates across three California counties. Combined with a queuing theoretic model, these results characterize the dynamic supervision requirements and thereby scalability of the teleoperation approach. Across all scenarios we find reductions in operator requirements greater than 99% as compared to in-vehicle supervisors for the time period analyzed. The work also demonstrates two methods for reducing these empirical supervision requirements: (i) the use of cooperative connected AVs -- which are shown to produce an average 3.67 orders-of-magnitude system reliability improvement across the scenarios studied -- and (ii) aggregation across larger regions.

Figures (6)

• Figure 1: Visualizations of each county and its OpenStreetMap interstate network overlaid in black.
• Figure 2: A map showing counties in southern California. The three selected counties are indicated with stars. Their diverse geographic areas and populations make them interesting study cases, and their proximity to each other suits the supervisor aggregation analysis.
• Figure 3: As lane-level detail cannot be seen in \ref{['fig:osm_nets']}, this figure shows greater detail for a subsection of the Los Angeles County traffic network reconstruction, as well as a lane-level pop-out.
• Figure 4: County supervision requirements with a 25% UCAV penetration rate. The maximum number of supervisors required during each 15-minute interval is plotted. This can be unpredictable; note LA County's afternoon spike.
• Figure 5: Supervision aggregation benefits: the gold line shows the sum total of supervisors needed across the three counties to achieve six 'nines' of reliability when supervision occurs on a per-county basis. The purple line shows the number needed when supervision tasks are pooled across counties.