The majority of self-driving cars that are currently being tested on the roads do so in a safe setting under the watchful eye of a human driver in good road and environmental conditions. According to research, there will be about 8 million automated or driverless automobiles on the road by 2025 [1]. Self-driving cars will need to develop through six layers of improvements in driver support technology before they can merge into roads.
Level 0: No Driving Automation
The majority of automobiles on the road today operate manually and are at this level. The dynamic driving task is performed by humans, and there are systems in place to support the driver. An information fusion technique is used to combine data from several sources, such as GNSS, vehicle motion sensors, and road maps. The human driver receives such map-matched information, together with information on the state of the traffic, for aid and direction. The emergency braking system is an illustration of Level 0 automation.
Level 1: Driver Assistance
The car has a single automated system for driver assistance, such as steering, accelerating, or cruise control, since it has the lowest level of automation. Adaptive cruise control, which allows the car to maintain a safe distance from other nearby vehicles.
Level 2: Partial Driving Automation
Advanced Driver Assistance Systems (ADAS) equipped vehicles have steering and acceleration/deceleration controls. At this level, a human driver occupies the steering wheel and is always in control of the vehicle. The Tesla Autopilot and Cadillac (General Motors) Super Cruise systems are two examples of this level of automation.
The computation of absolute localization uses GNSS, PPP or RTK, IMU, and odometry. To determine relative localization, perception sensors like cameras, radar, and/or LiDAR are employed. For steering control, acceleration, and brake control, both absolute and relative localization outputs are combined with path planning.
Level 3: Conditional Driving Automation
Level 3 vehicles are capable of detecting their environment and making intelligent judgments, such as assisting in an emergency stop by slowing down and stopping the automobile while sounding the horns of other vehicles. At this level, human override is still necessary. If the system is unable to complete the task, the driver must remain vigilant and prepared to take over.
A great example of Level 3 automation is the traffic jam pilot tool. The driver can sit back and unwind as the system manages all acceleration, steering, and braking. Real-time high-precision GNSS absolute positioning solutions are necessary to achieve lane-determination with or without a local base station network starting at Level 2 and higher. Many autonomous driving functions are unlocked by sub-lane level accuracy, which also allows for planning beyond perception.
Traffic Congestion: The Cost of Complacency
Roadways across the United States are plagued by congestion. According to an article in The New Yorker, approximately “one out of every six American workers commutes more than forty-five minutes, each way.” Additionally, 3.5 million American workers commute 90 minutes or more. With the number of motorists on the road and the number of miles traveled increasing every year, congestion has become a national problem in the United States.
Of course, many of these heavily congested roads rely on traffic signals to manage traffic. Yet 70-90 percent of traffic signals in the U.S. use outdated technology, are poorly timed and cause unnecessary delay. Day after day, Americans fight their way through an increasingly larger sea of vehicles and an abundance of red lights. Congestion and poor traffic control not only cause frustration for motorists and traffic professionals, but significantly diminish our quality of life, our health, and the wellbeing of our communities.
Congestion and poor traffic control costs individual motorists and American society-at-large millions of dollars in lost time, wasted fuel, environmental damage, and traffic accidents. In 2017, congestion cost American motorists $166 billion in wasted time and fuel. On an individual level, time wasted at traffic signals costs each motorist $1080 per year. Taking into consideration all transportation-related expenses, families now spend more money on transportation each year than they do on groceries.
The fuel wasted idling at red lights or in stop-and-go traffic also increases harmful emissions in the atmosphere, increasing air pollution and accelerating climate change. Over half of all Americans now live in areas with unhealthy levels of air pollution. Furthermore, studies have shown that regular exposure to air pollution increases the occurrence of asthma, cancer, lung and cardiovascular diseases and the related fatalities from these illnesses. Poor signal timing and congestion also compromise roadway safety. Every year, there are over 30,000 traffic fatalities with approximately 40% of crashes occurring at intersections. According to AAA, car accidents cost the US population $299.5 billion annually, or more than $1,522 for every man, woman and child in the US.
