• Visit our partners: Our Partners:
  • Visit our partners: Our Partners:

Driverless Vehicles: An Incredibly Close Future

Written by Dave Page

When we think of driverless vehicles, we often think of highly advanced machines such as those currently under development by Google, Uber, and Tesla. However, the idea of driverless vehicles can be traced back as far as the 16th century when Leonardo da Vinci first sketched his idea for a cart powered by coiled springs. Although it is not clear whether or not da Vinci ever built this device, scientists in Italy were able to create a working model in 2006 using da Vinci’s original plans. The cart was designed to be completely autonomous and was capable of completing simple circuits assuming that the steering had been pre-set in advance. Although this contraption seems little more than a toy when compared to what is available today, it is often described as not only the first autonomous vehicle but also the first practical example of a robot.

https://commons.wikimedia.org/wiki/File:Driving_Google_Self-Driving_Car.jpg

So, how did we get from this essentially clockwork powered cart to the truly incredible autonomous vehicles that are just starting to appear on roads around the world?

Today we will not only answer that question but also take a look into some of the fascinating moral and legal dilemmas that increased use of this technology present to the world at large. So, please join us as we take a look into the past, present and future of autonomous vehicles. 

The public were first introduced to the concept of driverless vehicles at the 1939 World’s Fair when, while at an exhibit by General Motors, he demonstrated that a car fitted with magnetic sensors would be capable of safely travelling along a specially modified roadway with magnetised spikes concealed beneath its surface. 1953 would see a slightly adapted version of this technology become a reality when RCA labs developed a miniature car that was able to drive by itself by detecting a series of wires embedded in the floor of a laboratory. 

This breakthrough inspired two employees of the Nebraska Department of Roads (Leland Hancock and L. N. Ress) to build a full-scale version of this system on an actual highway. This system was successfully installed at an intersection outside of Lincoln, Nebraska and, according to one report, it worked by “Burying a series of detector circuits in the pavement and pairing them with a series of lights alongside the road. The detector circuits sent impulses that guided the cars and determined the presence and velocity of the experimental cars on the road.” This experiment would prove to be hugely successful and on June 5th, 1960, the system would be demonstrated again In Princeton, New Jersey, at the headquarters of RCA labs.

However, it was not just the United States that were making progress with this technology. During the 1960s, in the United Kingdom, a modified version of the Citroen DS would set a new record by maintaining a steady speed of 80 miles an hour (129 km/h) around a test track in various weather conditions. This car, equipped with special sensors, would follow an electrical cable buried in the road and was consistently able to outperform human drivers.

Although most of the research into this style of autonomous vehicle appeared to be promising, there were two major drawbacks. These vehicles were only capable of driving on heavily modified roads which made them all but useless with the existing infrastructure. Also, the costs of modifying every single existing road would be so astronomical that no country seriously considered the idea, and scientists and car manufacturers would set about developing new technologies that would allow these vehicles to travel safely on ordinary roads. 

During the mid-1970s, scientists would develop technology that would allow small robots to follow lines using cameras and avoid obstacles with the use of sonar and, in the late 1970s, researchers working at the University of Tsukuba’s Mechanical Engineering Lab would use variations of this technology to carry out testing on Japanese roads, with the first autonomous vehicle that did not require any external cables or sensors in the Tarmac.

1984 would prove to be an absolute game changer for the advancement of autonomous vehicles as it would herald the beginning of Carnegie Mellon Universities’ NavLab project. Funded by the U. S. Department of Defence’s Advanced Research Projects Agency (ARPA) its mission was: “To use computer vision to create autonomous navigation” and, due to the advances in computer processing power, by 1986 they had developed the world’s first self-contained, fully autonomous vehicle and dubbed it NavLab 1.

https://commons.wikimedia.org/wiki/File:Navlab-1-5-autonomous-vehicles.jpg
Navlab models 1 (farthest) through 5 (front)

At an estimated cost of $1 million, the extensively modified blue Chevy van contained computers, cameras, several generators (at least one of which was required to power the air conditioning system that prevented the computers from overheating), and an early version of a laser rangefinder not dissimilar to the LIDAR systems employed by modern day autonomous vehicles.

By the time its replacement came along in the form of NavLab 2, a repurposed army ambulance Humvee, NavLab 1 had successfully achieved self-driving speeds of 12.5 mph (or 20 km/h).

Since the beginning of the project, there have been 11 NavLab vehicles, and each version pioneered new technologies that are used in the modern-day counterparts such as self-learning neural networks, accurate global positioning systems, and the previously mentioned LIDAR technology. The NavLab team demonstrated just how far the project had come when, in 1995, NavLab 5 made the 3,100-mile journey from Pittsburgh to San Diego with 98.2% of the trip being handled by the on-board systems alone.

In 1999, autonomous transport would take a huge step forward with the introduction of the ParkShuttle which, according to its website “is an electrically-driven, autonomous shuttle service that runs between Kralingse Zoom metro station in Rotterdam to the Rivium business park in Capelle aan den Ijssel”. Although the progress of this vehicle is overseen from the control room and it is possible for an operator to remotely start or stop the bus, the on-board computer system is fully capable of navigating by detecting sensors buried in the road and using additional radar sensors to detect any unexpected obstacles. Since its introduction, it has undergone several upgrades and now carries hundreds of passengers every day.

In 2002, DARPA announced the first Grand Challenge in which they offered a $1 million prize to anybody who was capable of constructing a fully autonomous vehicle that could navigate an extremely challenging 142-mile (228.5km) course through the Mojave Desert. Unfortunately, out of the 20 participants, none made it across the finish line. In fact, the winning entry only managed 8 miles (12.8km) before catching fire. This appears to have been a genuinely damaging outcome and, after the news of the fire was reported, many people seriously believed that fully autonomous vehicles were an impossibility. However, the second challenge, held in 2005, would see the deployment of far more sophisticated technology, and as a result of this 5 of the starting participants would successfully navigate the entire course.

The success of this project, at least in part, lead to many larger car manufacturers and tech firms to start their own independent research into autonomous vehicles and, it was at this point that the race for the world’s first truly independent driverless vehicle began.

With companies such as Google and Uber racking up serious test miles on public roads, concerns began to arise as to the safety, efficacy, and legal responsibilities that such vehicles pose. In 2018, one of Uber’s self-driving cars struck and killed 49-year-old Elaine Herzberg in Tempe, Arizona. This incident sparked a massive investigation as to who was at fault and the following is a brief outline of the results. According to an article by Andrew J. Hawkins:

https://commons.wikimedia.org/wiki/File:Uber_autonomous_vehicle_prototype_testing_in_San_Francisco.jpg

“When the board read aloud its findings on the probable cause of the crash in Tempe, the first person to be blamed was Rafaela Vasquez, the safety driver in the vehicle at the time of the crash. Vasquez was never called out by name, but her failures as a watchdog for the automated driving system were put on stark display by the NTSB.”

“In the minutes before impact, Vasquez was reportedly streaming an episode of The Voice on her phone, which is in violation of Uber’s policy banning phone use. In fact, investigators determined that she had been glancing down at her phone and away from the road for over a third of the total time she had been in the car up until the moment of the crash.”

“Vasquez’s failure “to monitor the driving environment and the operation of the automated driving system because she was visually distracted throughout the trip by her personal cell phone” was cited as the primary cause of the crash. But she shares the blame with her employers at Uber, where a woefully inadequate safety culture also contributed to Herzberg’s death.”

All this seems to be fairly straightforward, with the ultimate responsibility lying with the human in the driving seat. but what happens when the driver is completely removed from the equation? Companies such as Google are currently testing vehicles that have no controls at all, and these machines present an entirely different set of questions and problems.

One of the ideas that has been proposed is to include some variation of Asimov’s three laws of robotics but, this solution is not without problems of its own.

Let’s take a look at the first law: “A robot may not injure a human being or, through inaction, allow a human being to come to harm.” On the face of it, this sounds like a perfectly sensible thing that should be included in the route programming of every single driverless car on the road. Unfortunately, the staggering number of unknown variables presented by day-to-day driving makes this law impossible to follow. Consider, if you will, the following example:

A driverless vehicle is travelling at 30 miles an hour along a road when a small child steps out from between two parked cars 30 feet ahead. Even taking into account the on-board computer’s far superior reaction times, there is no way that the vehicle can come to a complete halt before striking the child. In this situation, the vehicle is presented with a number of options.

One.

Apply the brakes and hope that the child will sustain far less injury by being hit at a considerably lower speed.

Two.

Attempt to swerve around the child, potentially endangering other motorists.

Or three. 

Elect to voluntarily collide with one of the parked cars and risk injuring its own passenger.

As we can see, neither of these options are ideal and all of them have the possibility to break the first rule.

The second rule, “A robot must obey the orders given it by human beings except where such orders would conflict with the First Law” presents fewer issues, but only if we assume that the control system possesses better ethical judgement than the human travelling inside the vehicle. Elaborating on this, a human passenger may elect to risk harming themselves by colliding with a parked car or a lamp post rather than being involved in a head-on collision with a bus full of passengers – the second law in its current form would not allow for this decision to be made.

https://commons.wikimedia.org/wiki/File:Self_driving_vehicle.jpg

Another argument against the mass deployment of driverless vehicles is one that has been around for a long time: How will it impact on the jobs market? And this too, is an interesting question. How many jobs will be lost if taxi drivers are no longer required, or if fleets of driverless trucks are rolled out by haulage firms?

We are already reaching the stage where driverless vehicles are safer than human controlled vehicles, and they come with the added bonus to industry that they do not need to stop for toilet breaks, rest, or food. However, we are by no means at the point where we have solved the driverless vehicle challenge. According to William “Red” Whittaker, a CMU professor who led the development of the original NavLab project: “Of course, it isn’t solved, the kinds of things that aren’t solved are the edge cases.”

https://commons.wikimedia.org/wiki/File:Google_self_driving_car_at_the_Googleplex.jpg

And there are plenty of edge cases to contend with, including sensors being blinded or impaired by bad weather, bright sunlight, or obstructions. Then there are the inevitable software and hardware failures. But more important, the edge cases involve dealing with the unknown. You can’t program a car for every imaginable situation, so at some stage, you have to trust that it will cope with just about anything that’s thrown at it using whatever intelligence it has. And it’s hard to be confident about that, especially when even the smallest misunderstanding, like mistaking a paper bag for a large rock, could lead a car to do something unnecessarily dangerous.

In spite of these concerns, more and more autonomous vehicles are finding their way onto the world’s roads and with places like the United Kingdom already proposing changes to the highway code in order to better prepare for their arrival, coupled with the fact that   companies such as Google, Tesla, Uber, and most recently Apple, are pouring billions of dollars into research, it really does look like driverless vehicles are here to stay.

Related Articles

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Stay Connected

50,638FollowersFollow
829,000SubscribersSubscribe

Random Article

00:17:54

The Need for Speed: The Grumman F-14 Tomcat

Before jumping in to today’s video, let's all “come clean” by admitting that at some point in the past 3 ½ decades, each and...

Latest Articles