360 458 iStock 187178116

March 2017

Three perspectives on the latest advances in automated vehicle technology


iStock 000005913775Medium 1Breaking Down the Levels of Automated Driving

The Society of Automotive Engineers (SAE International) has established six levels of driving automation (SAE standard J3016), which the National Highway Traffic Safety Administration (NHTSA) has also adopted. Let me break them down for you.

SAE Levels are based on how much of the driving task is performed by the vehicle versus the human, based on control of speed/steering, monitoring of the environment and readiness to take over.

Level 0 is no automation; full human control. You’re talking about a vehicle that has no automated function, only the ability to enhance driver actions. It might still have assist features like cruise control, but it can’t monitor nor respond to outside events (like a stopped vehicle ahead) without human intervention.

Level 1 introduces driver assistance systems. When you use a driver assistance system like adaptive cruise control (ACC), the technology maintains the set speed and distance from the car ahead of you instead of the driver, so you don’t need to worry about slowing down if the car ahead of you slows down (within the capabilities of the system). These assistance systems can also include active safety systems, like automatic emergency braking (AEB), which can detect when you need to brake and even brake for you.

Level 2 is partial automation, which gives the vehicle responsibility for more advanced tasks but requires a human to always be monitoring and in control. An example of this is adaptive cruise control combined with a lane-keeping assist system. The technology will automatically adjust your speed as well as keep you in the designated lane. However, the systems have some dependencies; for example, they rely on lane markings and can’t function if there aren’t any, so the human must remain in control.

Level 3 is conditional automation. This increases the responsibilities of the vehicle for execution and monitoring. Under certain conditions, the car is functioning in an automated way and almost has full control. It’s like an aircraft autopilot, controlling your speed and steering, keeping you in your lane, monitoring your blind spots and traveling a navigated route. However, these systems still have some dependencies (like needing lane lines); therefore, the driver needs to be ready to take over at all times if the conditions aren’t conducive to the technology functioning properly.

In levels 4 (high automation) and 5 (full automation), the vehicle has complete control and the human is not responsible for the trip. The differentiator between levels 4 and 5 is that level 4 vehicles can only operate in certain design conditions, like a limited-access highway corridor or closed campus. Level 5 vehicles could operate anywhere.

Almost everyone immersed in the autonomous vehicle field agrees that this is going to be an iterative implementation of technology. Even when the technology is fully ready, there are going to be places that aren’t going to have the necessary infrastructure for level 5 vehicles to operate, from a safety perspective or otherwise.

The potential of these systems to enhance safety is providing the industry with critical motivation. There have been a number of studies, including many conducted by the Insurance Institute for Highway Safety (IIHS), that show the effectiveness of active safety systems in reducing the frequency and severity of insurance claims. As a result, because systems like AEB show so much promise and we have the data to show the effectiveness, manufacturers have come together and committed to standardization.

One of the biggest challenges for the implementation of automated driving systems on our roadways is that we need customers to feel comfortable with the technology and understand it. There’s the danger of people being so in awe of the incredible abilities of a level 2 or 3 vehicle, for example, that they may think the vehicle is more capable than it is. We don’t want people thinking the vehicle can do more than it actually can. Right now, humans need to be ready to take over. It’s important that we keep the public educated so they understand the technology as it evolves.

iStock 000005913775Medium 2Automated Vehicles in Pennsylvania

Technology is changing the world every day, and there are few places where that change is more evident than inside an automobile. We are closer to the possibility of fully automated vehicles than we have ever been. Before these vehicles can travel on our highways, however, they need to be thoroughly and safely tested. Pennsylvania has become a hub of innovation related to highly automated vehicle (HAV) technology. Pennsylvania is home to acclaimed institutes of higher learning, such as Carnegie Mellon University and the University of Pennsylvania, that have been trailblazers in automated vehicle research. Pennsylvania has a varied topography and weather patterns, making real-world HAV operation attractive to entities that want to test their technology. This fact, combined with the academic resources there, has created a fertile testing ground for automated vehicles by companies such as Uber.

Recognizing PennDOT’s role as the lead agency in the commonwealth, we established an Autonomous Vehicle Policy Task Force, which began meeting in June 2016. The task force has a clear goal: to develop policies to expand the testing of automated technology in Pennsylvania while ensuring public safety.

The task force comprises industry leaders, academic experts and constituent representative groups from both the public and private sectors. The task force focuses on the following areas: setting minimum approval criteria for testing, vehicle characteristics and capabilities, the conditions under which testing can occur, data collection, testing approval and renewal, and the question of “Who is the driver?” The task force has utilized the recently released National Highway Traffic Safety Administration (NHTSA) automated vehicle policy guidance document—especially as it relates to the Model State Policy—as a framework for addressing areas of concern (notably on-road testing).

Recently, the task force submitted its draft policy recommendations to Secretary of Transportation Leslie Richards. Understanding that regulations, by design, are a slow process, PennDOT identified early in the process that establishing policies would offer the maximum amount of flexibility in addressing the rapidly changing HAV technology advances. Working in collaboration with the Pennsylvania legislature, SB 1412 was introduced. The legislation, when enacted, will enable PennDOT to implement identified policies for HAV testing programs. The legislation is being refined by the General Assembly in consultation with the task force.

PennDOT held a webinar in December 2016 to collect public comments related to the proposed policies. Watch a recording of the webinar.

iStock 000005913775Medium 3aA Road to Reliable Transportation

Self-driving technology is an extension of Uber’s mission to provide transportation that’s as reliable as running water everywhere, for everyone. We set out in 2010 to solve a simple problem: How do you get a ride at the touch of a button? At the time, the idea was unthinkable. Nearly seven years and over two billion trips later, it’s a reality in more than 460 cities in over 70 countries around the world.

Self Driving Uber
Self-Driving Ubers are being tested on the roads of Pittsburgh. During testing, there is a safety driver in the front seat because the vehicles require human intervention in many conditions, including bad weather.

Almost two years ago, we launched our Advanced Technologies Group in Pittsburgh with a goal of creating our Self-Driving Ubers, which made their debut this past September in the Steel City. This pilot is a big step forward. Real-world testing on Pittsburgh’s challenging roads in its varied weather is critical to our success. We’re pleased with the city’s enthusiastic reception to our pilot, as well as the progress we’ve made since launching.

We can imagine how this technology could further extend the mobility benefits people already enjoy thanks to ridesharing. By getting more people into fewer cars, we can help reduce congestion and pollution over time. In Los Angeles, UberPOOL reduced the number of miles driven by 7.9 million, and the amount of carbon dioxide emitted by 1,400 cubic tons in its first seven months.

We’ve also seen how Uber increases mobility for everyone, particularly for people living in underserved areas. Take Manhattan, where 35 percent of all Uber pickups happen outside of the borough, compared to just 6 percent by traditional yellow taxis.

But nothing is more inspiring than the tremendous potential self-driving technology holds to improve road safety. Today, 1.3 million people around the world die in car crashes every year. Ninety-four percent of those crashes involve human error. This is a tragedy innovation can help solve.

A better future is within reach. We have the technology. While it won’t happen overnight, self-driving will be an important part of the future of transportation.