CONNECTED & AUTONOMOUS VEHICLES


THE LONG ROAD TRAVELLED:
A self-driving history

Many believe that the development of self-driving vehicles is a particularly 21st century phenomenon. In fact, their development goes back much further, starting in the 1920s. The first proto-vehicles appeared in the 1980s, and numerous multinational car manufacturers have taken the quest forward since then. Rich Porter looks at self-driving vehicles’ long history and what lies ahead for this promising and disruptive technology

It all began in 1925 when Houdina Radio Control demonstrated the 'American Wonder' in New York. This vehicle was a 1926 Chandler fitted with a transmitting antenna, which received radio impulses from another car following behind.

Leap forward into the 1930s and American theatrical and industrial engineer Norman Bel Geddes demonstrated an early representation of an automated car, in a General Motors sponsored exhibit, at the 1939 World's Fair. This demonstration showed a radio-controlled electric car propelled via electromagnetic fields from circuits embedded in the highway. After the Second World War RCA Labs took up the baton in 1953 when they built a miniature vehicle that could be controlled via wires laid out on a laboratory floor.


This trial inspired Nebraska Department of Roads traffic engineer Leland M Hancock and his Director, LN Ress, to experiment with this system in a highway environment. In 1957 a full-size system was demonstrated by RCA Labs and the State of Nebraska on a 400-foot strip of public highway.

In the early 1960s the UK government was also working on then cutting-edge technology. The Transport and Road Research Laboratory tested a driverless Citroen DS on a specially prepared stretch of the M4 motorway. The car interacted with road-based magnetic cables and was capable of 80 miles per hour without any driver inputs to the steering, brakes or accelerator. The test vehicle is today preserved by the Science Museum in London as the world’s first completely automatically controlled road vehicle.


Tracking forward further still and the 1980s saw German automotive giant Mercedes Benz develop

a vision-guided robotic van. Designed by Ernst Dickmanns and his team at the Bundeswehr University Munich, the van notched up speeds of 39 miles per hour on traffic-free streets.

Houdina Radio Control demonstrated the 'American Wonder' in New York in 1925

Tracking forward further still and the 1980s saw German automotive giant Mercedes Benz develop

a vision-guided robotic van. Designed by Ernst Dickmanns and his team at the Bundeswehr University Munich, the van notched up speeds of 39 miles per hour on traffic-free streets. The 1980s also saw the lynchpin Defense Advanced Research Projects Agency (DARPA)-funded Autonomous Land driven Vehicle (ALV) project in the US. This competition drove many cutting-edge technologies, including the first road-following demonstration leveraging lidar, computer vision and autonomous robotic control to direct a robotic vehicle at speeds of up to 19 miles per hour. The pace of self-driving development accelerated even further into the 1990s. In 1991 the US Congress passed legislation to "demonstrate an automated vehicle and highway system by 1997".


The Federal Highway Administration (FHWA) seized the challenge with a series of initial analyses, then establishing the National Automated Highway System Consortium (NAHSC). This project was led by FHWA and General Motors but featured several other additional partners. The work of this project bore Demo '97 on I-15 in San Diego, California, which saw 20 automated vehicles put through their paces.

“In 1994 twin robot vehicles, VaMP and Vita-2, were developed by Daimler-Benz and Ernst Dickmanns. These vehicles drove over 620 miles on a three-lane highway in standard heavy traffic in Paris at speeds up to 81 miles per hour, albeit semi-autonomously”

In 1994 twin robot vehicles, VaMP and Vita-2, were developed by Daimler-Benz and Ernst Dickmanns. These vehicles drove over 620 miles on a three-lane highway in standard heavy traffic in Paris at speeds up to 81 miles per hour, albeit semi-autonomously with occasional interventions. Also, in 1994, Lucas Industries developed parts for a semi-autonomous car in a project funded by the UK Department of Trade and Industry, Jaguar Cars and Lucas.

Further breakthroughs occurred in 1995 and 1996 via projects from Carnegie Mellon University (Navlab), Ernst Dickmanns (re-engineered autonomous S-Class Mercedes-Benz) and the University of Parma (ARGO Project). To round off the decade two pilot projects took place in the Netherlands, at Schiphol Airport in 1997 and at business park Rivium in 1999. In these trials ParkShuttle, an automated people mover leveraging magnets in the road as reference points, carried passengers, making the claim that is was the first driverless vehicle – as it had no steering wheel, pedals or safety drivers/stewards on board.

The start of the new millennium saw the US government pour funding into several military projects centred around the development of unmanned ground vehicles, which would be able to navigate difficult terrain and avoid trees and boulders. Jump forward to 2004 and DARPA launched the initial Grand Challenge. This milestone competition offered a prize of US$1 million to any engineers capable of developing an autonomous vehicle that could finish a 150-mile course in the Mojave Desert. While no team could achieve this, another Grand Challenge was held in 2005. That year five vehicles managed to complete a course with GPS points and obstacles. The third event was held in 2007 with an autonomous Chevy Tahoe from Carnegie Mellon University taking first place. The DARPA events were a huge boost for research into self-driving vehicles and how they could drive down congestion and reduce traffic accidents. In 2009 technology giant Google made the key decision to develop autonomous cars but did not announce this publicly. As of 2013, many of the automotive giants, such as BMW, Volvo, Audi. Nissan and General Motors, among others, were testing autonomous cars.

From 2010 onwards, self-driving development activity ramped up steadily, with the University of Parma running the VisLab Intercontinental Autonomous Challenge. This 9,900-mile test run was the first intercontinental land journey completed by autonomous vehicles. The vehicles travelled from Parma, Italy, to the Shanghai Expo in China over 100 days. In Germany, the Institute of Control Engineering of the Technische Universität Braunschweig demonstrated the first autonomous driving on German public streets. Over in the US, Nevada passed a law on the operation of autonomous cars in the state in 2012. That same year Nevada motor vehicle examiners administered a 14-mile driving test by a Google self-driving car, which it successfully passed. The first license issued to a self-driving car in the US was a Toyota Prius modified by Google's experimental driverless technology in Nevada. Florida became the second state in America to permit the testing of these vehicles on public roads, followed by California.

The following year, in 2013, there was a huge amount of development activity, passenger vehicle tests in the Netherlands, a further DARPA Grand Challenge in the US and Daimler R&D, with Karlsruhe Institute of Technology/FZI, developing a Mercedes-Benz S-class vehicle which drove autonomously for about 100km. Japanese giant Nissan also announced plans to launch a range of self-driving cars by 2020 and began work on a proving ground for said vehicles – which was completed in 2014. A Nissan Leaf electric vehicle was fitted with its autonomous driving technology and demonstrated at an event in California; later being granted a license plate to enable it to drive on Japanese public roads. Although 2013 saw a lot of testing activity, none of the above vehicles were accessible to the public yet.

“The UK is at the forefront of not just public road trials but also in pushing the boundaries of what social benefits the technology can bring about.”

However, that changed when the Navia Shuttle, developed by Induct Technology, became the first self-driving vehicle for sale commercially in 2014. Resembling a golf cart and with a top speed of 12.5 miles per hour, it seated eight and was designed to be used in pedestrianised urban areas, industrial sites, airports, hospitals or theme parks. That same year Google ramped up its activity with an announcement it would launch 100 autonomous car prototypes developed in its secret X lab. Towards the end of the year, Tesla announced the first AutoPilot. Its Model S cars fitted with this system were capable of lane control with autonomous steering, braking and speed limit adjustment based on signals image recognition. The system also offered autonomous parking and could improve over time via software updates. In early 2015, the UK government unveiled public trials of the LUTZ Pathfinder driverless pod in Milton Keynes, followed by Tesla announcing the introduction of its Autopilot technology by the midpoint of the year; via a software update for the cars equipped with systems to enable autonomous driving.


In 2018 British firm Oxbotica announced that it had entered into a strategic alliance with minicab fleet Addison Lee to develop robo-taxis and just last month, April 2019, another British company, FiveAI, began trialling its self-driving Ford Mondeos on the roads of the London Boroughs of Bromley and Croydon. The UK is at the forefront of not just public road trials but also in pushing the boundaries of what social benefits the technology can bring about. Aurrigo is an autonomous pod manufacturer and it is currently working with Blind Veterans UK to test last-mile solutions for the visually impaired in Brighton

The work going on today shows how self-driving vehicles could trigger huge social, industrial and economic benefits. The market has the potential to be worth £51 billion to just the UK economy by 2030, according to the Society of Motor Manufacturers and Traders (SMMT). The development of these vehicles could expand the industrial base, enhance productivity, boost road safety levels, improve road congestion and reduce pollution.


Europe is making great strides with several countries legalising testing on public roads, including France, Germany, the Netherlands, Norway, Sweden and the UK. Germany, the Netherlands and the UK have all reviewed legal frameworks, comprising licence equivalents for self-driving vehicles. The momentum is also growing in the US, with California passing state level approval for driverless vehicle testing with no safety driver present; while the US Department of Transportation issued guidance for automated driving pilot programmes. China added 11 roads to the existing 33 in Beijing for autonomous driving tests, while Audi, BMW and Daimler were permitted to test vehicles in Beijing and Shanghai. However, the UK is the country leading the charge and has been instrumental in the development of EU safety and testing legislation.


There are still numerous challenges ahead, not least pertaining to the negative perceptions among the public around safety. These issues will need to be faced head-on if the opportunities currently on offer are to be seized. Safety is a major focus and many governments around the globe appreciate this; none more so than the UK. The UK government’s Centre for Connected and Autonomous Vehicles’ (CCAV) updated its Code of Practice to provision for self-driving vehicles to operate on the country’s roads. The code has made allowances for a remote safety operator to take control of the vehicle, if there is a problem. Any tests must be conducted in the utmost safety, because accidents will simply result in further negative perceptions, which could impinge on the success of projects. The updated code also mandates that manufacturers, and any parties wanting to undertake trials, must publish a safety test case. This document needs to contain any pertinent information for the trials, such as the vehicle being used, what the trial comprises, evidence that it can be conducted safely and what the interactions with the authorities will be, among others.

This information is extremely valuable in establishing an increasingly methodical plan for safe self-driving development. It also underlines the crucial nature that codes of practice and legislation play in this journey. They are vital to establishing uniform procedures and expectations around responsible trials. All relevant stakeholders must work together to connect policy and technology to boost safety. Alongside safety, cyber security is also a concern. The systems controlling self-driving vehicles, and the vehicles themselves, are all at risk from cyber-attacks. When testing and developing the infrastructure around these vehicles, stakeholders must focus on cyber security. This requires the development of water-tight cyber security standards – to boost readiness and resilience to attacks. Progressive standards and codes of practice are the only way to ensure this happens. At the centre of these protocols is the need to continuously share information. The most vital area for collaboration of all is the sharing of data created by all aspects of the ecosystem. Trials will deliver extremely valuable data, which can be successfully harnessed for greater good. For example, data shared with the authorities on the performance of self-driving vehicles on the road will help to avoid problems in the future.


Meeting these challenges now is only possible if all the stakeholders work together to ensure that supporting legislation moves in sequence with the development of technology.


Cracking this will be the winning formula in the challenge to champion self-driving vehicle development.

FYI


Rich Porter is Innovation and Technology Director, Zenzic, formerly Meridian

Share now...