ATSSA attends 2018 Automated Vehicle Symposium

ATSSA recently co-moderated a session entitled, “Infrastructure Readiness, Reading the Road Ahead,” that addressed the ways in which U.S. infrastructure will best accommodate Connected Automated Vehicles (CAVs) during the 2018 Automated Vehicle Symposium, which was held in San Francisco, California, in early July.

The session was broken out into two different topic areas: the first being how roadway infrastructure is read by CAV systems. Moving forward, roadways might require a classification system to alert CAVs of the Levels of Automation (LoA) they can support. LoAs can range from one to five, with one being no automation and five being full automation.

During “Infrastructure Readiness, Reading the Road Ahead,” Transpo Group’s Director of Safety and Mobility Solutions Scott Kuznicki stressed the importance of consistent use and the standardization of Traffic Control Devices (TCDs). Variations of TCDs from state to state would lead for inconsistent support of CAVs and the Advanced Driver Assistance Systems (ADAS), which allow CAVs to operate autonomously through several machine vision sensor systems.

Kuznicki furthered the discussion on road classification systems to ensure AVs are driving on roads that can accommodate them. The secondary overarching topic related to various issues and possible failure modes that machine vision systems encountered. The failures of machine vision systems could be caused by a number of instances including inconsistencies in roadway infrastructure or blockages of the lidar sensor, which prohibits the sensor from scanning the roadway.

ADAS rely heavily on Light Detection and Ranging (LIDAR), radar, and computer camera vision sensors to read the surrounds of CAVs and guide the vehicles on roadways. Leo A. McCloskey of Echodyne and Robert Seidl of Motus Ventures provided an overview of the sensors and explained that while these technologies are improving, they still have yet to provide full automation.

Additionally, several other critical issues and topics were addressed at AVS including:

• Standardization for next generation sensors and infrastructure;
• Optimizing infrastructure devices and structures to be more visible to machine vision sensor systems;
• Machine vision system failure modes, which included environmental sensor issues such as degraded or unclear pavement markings, degraded or hidden signage, and unsteady traffic lights;
• Digital infrastructure;
• Enabling technologies related to communications, sensors and sensor fusion (which is how camera, radar, and lidar sensors interact and determine which sensor is relied upon when there are conflicts between them), positioning, driver monitoring, and high-performance computing;
• Dedicated lanes for Connected Automated Vehicles (CAVs);
• Highway work zones;
• Data for transportation systems;
• The U.S. DOT Connected Work Zone Grant;
• Construction lane closures/work zone mapping for lane closures; and
• Smart work zones

During AVS, U.S. DOT Transportation Secretary Elaine Chao also reiterated that the agency’s top priority is safety and also addressed some of the challenges of vehicle automation related to technical aspects, ethics, liability, human behavior, perception, and acceptance.

AVS is produced in partnership with the Transportation Research Board (TRB). The 2019 event will be held in Orlando, Florida, Monday through Thursday, July 15-18.