Robotic Cars that will Drive us into the Future

Introduction

The idea of robot cars has always been a question of how and when. Back in the early ‘80s, the closest thing you had to an artificially intelligent car was Kitt, an electronically-aided car as a 1982 Pontiac Firebird Trans Am as seen on the famous American television series ‘Knight Rider’ with Michael Knight as a field agent saving the day in a black leather jacket and a hairy chest. Interestingly, unlike big hairy chests, the design and engineering of robotic cars may soon cruise into modern-day research and development practices to pave the way for a new concept to customer cars.

In 2010, Researchers at Stanford University demonstrated their design and engineering efforts by presenting the autonomously-driven car named Shelley. The research was led by mechanical engineering associate Professor Chris Gerdes. Take a look at the video below, it demonstrates the autonomous car ‘Shelley’ being taken through a practice workout on the grass at the Santa Clara County Fairgrounds in San Jose where she explores the racing tracks and turning laps.

Structure to an Autonomous Car

A vehicle designed to be completely autonomous will able to perform all of the tasks to move itself, tasks that would normally require a human. Not only should an autonomous vehicle be able to take full control of the vehicle, but great emphasis should also be placed on the autonomous car’s ability to sense its environment and manipulate the function and speed of the car accordingly to allow for accuracy in movement when adapting to varying driving conditions. The idea of putting an autonomous car ‘through the mill and back’ was tested in 2010 by Stanford’s Robotic Audi that drove itself along the Pikes Peak Race Course in Colorado.

To understand what’s involved in the design and engineering of a robotic car, it’s important to be aware of the components in the car that will make it react without being manually controlled by a human. Let’s sit back and take a look at an overview of some of the main actuation and sensory components in an autonomous vehicle that are fundamental in allowing a vehicle to sense its environment.

Actuation to a vehicle is possible with the application of direct current (DC) motors and servomotors. The DC motors are important for controlling movement of the rear wheels to a car and are key to converting an electrical current to torque (a product of force that results in the rotation of an object). The general concept here being that the more force applied to the DC motor, the more torque produced to power movement of the shaft. In the context of robotic cars, a digital signal would be required to power the DC motor using Pulse Width Modulation, but motor control cannot work via digital signals because it is difficult to deliver circuitry to the motor. However, an interface circuit called an H-bridge can help convert a digital signal that is of low power into a signal that is strong enough to isolate the DC motor and activate it.

This type of motor works to monitor the position and speed of a moving vehicle via a feedback loop to control the speed generated – a device used to provide precise control of motor output.

In order for the vehicle to adapt to its environment, there needs to be internal sensing systems in the vehicle, and there are. Thousands of years back, you’d have seen carts and waggons with the main sensor to control the movement and speed of the cart or waggon being you, the human. Luckily, modern-day technology has freed us all from pulling a waggon. Automated vehicles, are built with infrared sensors that can locate a white line on a black road surface, but the main issue with this type of sensor is that for them to accurately locate a white line on the road, the line has to be directly underneath the car; however, this is not the case for today’s road infrastructure – the white line is normally beside the car.

Magnetic sensors, such as Hall Effect Sensors isolate a magnetic link which activates a field effect device to then initiate an internal circuitry.

Vision sensors in the form of a digital camera positioned at the front end of a vehicle visually feed back the current state of the road in front. Using digital cameras in a vehicle brings the idea of autonomous cars one step closer to becoming more intuitive as visual feedback of the road ahead will resemble how the human brain perceives their environment when driving. Image processing whereby the vehicle offers information about the road ahead is only part of the image sensor system. It is also important to know how close another object on the road is to your car and this is sensed using headway sensors. The functional principle to headway sensing is based on the positioning of an object on the road (i.e., a moving object closer to the headway sensor will result in a more intense light emitted by the detector in this device that generates a voltage proportional to the distance between the vehicle and the object it has close contact with).

Robotic Cars Controlled by a Smartphone

The idea of robotic cars that can function without being controlled by a human is clearly an exciting step in automation design and engineering. Researchers are taking this idea of autonomous vehicles to another level by studying the application of smartphone technology to drive a car.

Griffith University robotic researcher Dr Jun Jo is close to finishing the design and engineering of a self-driving car that is controlled by a smartphone. Dr Jun Jo has described this technology as a vision-based system that responds to your commands via the smartphone. The technology in this futuristic car will include a lane detection system, a laser detection system, and ranging sensor to help identify and locate moving objects surrounding the car. The idea of a phone being able to control your car’s movement and speed makes me wonder about the level of safety with this technology. Regardless of smart vision-based systems to manipulate an automated vehicle, such technology may have technical breakdown issues and this is not what you need when driving along a busy road at high speed. Dr Jun Jo has address this issue of system malfunction and suggested that an external PC can also run car control and will work as a safety backup in the event of a malfunction to the smartphone.

Research efforts into building robotic cars are starting to look promising. Considering how I’ve never passed my driving test, the thought of robotic cars becoming mainstream at some point in the near future where I can just sit back and relax in a vehicle makes me wonder whether I ever need to book lessons with the AA Driving school.

References

• Kachroo, P., Mellodge, P. (2005). Mobile Robotic Car Design. USA, New York: The McGraw-Hill Companies, Inc.
• Professor J. Christian Gerdes. Stanford University.
• http://www.griffith.edu.au/create-whats-next/driverless-cars.