Towards Full Automated Drive in Urban Environments: A Demonstration in GoMentum Station, California

Each year, millions of motor vehicle traffic accidents all over the world\ncause a large number of fatalities, injuries and significant material loss.\nAutomated Driving (AD) has potential to drastically reduce such accidents. In\nthis work, we focus on the technical challenges that arise from AD in urban\nenvironments. We present the overall architecture of an AD system and describe\nin detail the perception and planning modules. The AD system, built on a\nmodified Acura RLX, was demonstrated in a course in GoMentum Station in\nCalifornia. We demonstrated autonomous handling of 4 scenarios: traffic lights,\ncross-traffic at intersections, construction zones and pedestrians. The AD\nvehicle displayed safe behavior and performed consistently in repeated\ndemonstrations with slight variations in conditions. Overall, we completed 44\nruns, encompassing 110km of automated driving with only 3 cases where the\ndriver intervened the control of the vehicle, mostly due to error in GPS\npositioning. Our demonstration showed that robust and consistent behavior in\nurban scenarios is possible, yet more investigation is necessary for full scale\nroll-out on public roads.\n