Are Self Driving Cars the New Future?

By: Victoria Hennemann

New Technology 

As time moves forwards thousands of technological advances are being researched and applied to everyday life. One of the most common technologies people use everyday are cars, however not everyone is able to drive. Those who have disabilities usually cannot drive. Currently the Americans with Disabilities Act of 1990 states transportation services must be offered to people with physical disabilities, visual or mental conditions, or injuries that prevent someone from driving on their own. However, paratransit which is taxi service can be expensive. Many researches have been working on a new technology that could potentially be a solution. This solution is a self driving vehicle. Srikanth Saripalli who was on the testing team at the Texas A&M campus reports that Autonomous shuttles can be a possible solution. Their goal is to create a fully integrated system which allows users to connect to the dispatching system and create profiles that contain the information about the passenger’s disability. They would also like the autonomous shuttle to have communication preferences and have specified destinations.

Picture 1. Autonomous Shuttle 

The Solution 

Many researches have been working on a new technology that could potentially be a solution. This solution is a self driving vehicle. Srikanth Saripalli who was on the testing team at the Texas A&M campus reports that Autonomous shuttles can be a possible solution. Their goal is to create a fully integrated system which allows users to connect to the dispatching system and create profiles that contain the information about the passenger’s disability. They would also like the autonomous shuttle to have communication preferences and have specified destinations.

Picture 2. Autonomous Shuttle Driving

Essentially what Saripalli and his team wants a person to be able to request a shuttle, then the system would find a vehicle that has the equipment the passenger needs. When the shuttle arrives to pick up the rider it would scan the area with lasers, cameras, and radar to create a 3D map of the area. Then it would use this information and merge it with traffic or information from Google Maps or similar navigation systems. This would help determine where the passenger needs to be picked up, it would identify an area that is easy to pick the passenger up, and send a message to the passenger when the shuttle arrives at the particular destination. The last goal is to have the vehicle actually communicate with the passenger about arrival time and potential detours. This would usually be displayed via text message or on a screen and the passenger would accept the typed input. Self driving cars are a new opportunity to give more affordable transportation access to everyone. The autonomous shuttle is one prototype that could potentially change people’s lives.

Picture 3. Srikanth Saripalli 

Interview with Srikanth Saripalli (click on link)

More Research

Jason Kong, Mark Pfeiffer, Georg Schildbach, and Francesco Borrelli conducted a kinematic versus dynamic vehicle model to see which one would be more effective when creating a design for a controller for an autonomous vehicle. The main components in the vehicle which Srikanth Saripalli also mentioned in his experiment are localization, perception, and control. So in this experiment that these four men conducted, they focused primarily on the control of the vehicle’s acceleration, brake, and steering using Model Predictive Control (MPC). Most published experiments for this type of research use dynamic vehicles models. The disadvantages to using this is that it can be expensive and the tire model becomes singular at lower speeds. Since the tire models use an angle estimation term which means the vehicle's velocity would be in the denominator when dividing. So creates a potential variable when designing the control design because it prohibits the use of stop-and-go scenarios. So Kong, Pfeiffer, Schildbach, and Borrelli proposed that a kinematic bicycle model be used. These two models were used to predict vehicle’s future states then this was compared to the measured states taken in the experiment. Then they used these results to develop a design for the controller of the autonomous vehicle using the MPC and the kinematic bicycle. The experiment showed that the kinematic bicycle model was most effective when using at different speeds on windy roads.

Sources: 

Kong, Jason, et al. "Kinematic and dynamic vehicle models for autonomous driving control 

     design." Intelligent Vehicles Symposium (IV), 2015 IEEE. IEEE, 2015.

Saripalli, Srikanth. “Are Self-Driving Cars the Future of Mobility for Disabled People?” Phys.org,
Science X Network, 6 Oct. 2017, phys.org/news/2017-10-self-driving-cars-future-mobility-
disabled.html.

Witte, Kathleen. “Focus at Four: Take a Ride with No Driver in A&M's Autonomous Vehicle.”

KBTX-TV, Gray Digital Media, 26 July 2017, www.kbtx.com/content/news/Focus-at-

Four-Take-a-ride-with-no-driver-in-AMs-autonomous-vehicle-436826053.html.