How long does it take for the driverless car to become a “normal” thing? According to analysts 5 years at most. The road test of Waymo, the spin-off of Google that works on the project, has been the first of these days and it has for the first time tested in Phoenix, Arizona, the circulation of driverless cars without any ‘human’ companion on board.
A turning point in the development of the driverless car, prior to its arrival on the market and its dissemination, up to its ‘normalization’, becoming a common use tool. The industry seems already worthing 7 trillion dollars, not surprisingly it tempts all the big automotive industries as well as tech companies, supported by technological know-how behind this revolution.
A mini “how to” brings together all there is to know about driverless cars, the current status, its key players, new industrial background.
BRIEF HISTORY OF THE DRIVERLESS CAR
From Tuscany, to the United States and Europe, the driverless car is a dream that many people have worked on, from worldwide. The first driverless car can be traced back (with a bit of flexibility and imagination) even to Leonardo Da Vinci. Around 1478 the brilliant Tuscan inventor and artist drew a wagon that could move without being pushed or pulled. The kinetic energy was supplied to the self-propelled wagon by a sophisticated system of toothed wheels set in motion by a pair of springs also providing the possibility to set in advance the path to follow. This mechanical basic robot is sometimes considered the ancestor of the modern car.
In 1920, the first experiments on automated machines were conducted. Some tests were carried out in the 1950s. In 1977 the Japanese Tsukuba Mechanical produced a driverless car capable to recognize the road, during the trip made at a speed of 20 miles per hour, thanks to two on-car cameras.
The first truly autonomous car appeared in the US in the 1980s: starting from 1984 NavLab of Carnegie Mellon University in Pittsburgh, Pennsylvania, realized a series of computer controlled vehicles for assisted and autonomous driving: car , trucks, SUVs and buses.
In 1987 the pan-European Eureka Prometheus (PROgraMme for a European Traffic of Higher Efficiency and Unprecedented Safety, 1987-1995) was launched, the largest research and development project in the field of driverless cars ever realized. The initiative received EUR 749 million in funding from Eureka Member States together with the contribution of several universities and car manufacturers. It shown event the collaboration with the German engineer Ernst Dickmanns, a pioneer of the driverless car, and with his team at the Bundeswehr Universität München, in partnership with Daimler-Benz.
In 1994, two twin robot vehicles, VaMp and VITA-2, traveled over 1000 kilometers on a Parisian motorway in traffic at a speed of 130 kilometers per hour.
The following year, in 1995, Ernst Dickmanns re-engineered a driverless Mercedes-Benz S-Class and drove 1000 miles from Munich to Copenhagen, Denmark, using computer vision.
Since then, several car manufacturers and research institutions have developed prototypes of driverless cars. These include: Mercedes-Benz, General Motors, Continental Automotive Systems, Autoliv Inc., Bosch, Nissan, Toyota, Audi, Volvo, Vislab, Parma University, Oxford University and Google. The Italian experiment of VisLab will be discussed below.
In 2013, four US states passed laws permitting traffic of driverless cars: Nevada, Florida, California, and Michigan. In Europe, some cities in Belgium, France, Italy and the UK are planning to activate transport systems for driverless cars. Germany, the Netherlands and Spain have already allowed some tests of this cars on the road.
DRIVERLESS CAR, THE TIMING
Driverless car technology is already available, but its adoption will depend on the “transportation as a service” developments, which will be focused on an overall mobility experience. Whether and when this experience will be implemented depends on governments, car manufacturers, software companies and all players working together in the new mobility system. Some countries may launch driverless cars faster than others. Problems related to the use of this vehicle are mainly linked to insurance, lack of regulatory standards and lack of infrastructure. The timing depends on all these factors. However, analysts and experts have already delivered their forecasts.
BY 2019 – Elon Musk, Tesla’s founder, has ensured that the fully driverless Tesla will arrive in 2019.
BY 2020 – “GM’s driverless cars – says Deutsche Bank analyst Rod Lache – will be ready for commercial deployment, without human drivers, much earlier than widely expected (within semesters, not years) and potentially years before competitors. We consider that business built on this platform will grow much faster than expected, and fast growth will provide substantial benefits”. According to Lache, General Motors would be in a particularly advanced position to be able to start mass production and monetize investments in this sector in the short term (within a year and a half).
BY 2020 – The first taxi-robots could be on the roads in 2020. At least that’s what is said by Jensen Huang, CEO of Nvidia, the Californian chip giant that develops technologies for driverless cars. This is not surprising: Nvidia is in fact the artificial intelligence used so far by many of the automotive industries. Powerful processors able to replace the man behind the wheel. Speaking in Tel Aviv at the Nvidia GPU Technology Conference, Huang said that as early as 2019 we will see several pilot projects, but 2020 will be the key year, tipping point in which the automated level 5 taxi cabs (the maximum classification of the American SAE) will come true.
BY 2022 – “Self driving cars could arrive in many markets within 5 years, their full adoption should take place in 10-15 years”, says an expert in the field, Randy J. Miller, EY Global Leader Automotive & Trasportation, in this interview with EconomyUp.
TESLA SELF DRIVING CAR
People who think of the self-driving car often think of Tesla, the company that produces electric cars founded by the visionary entrepreneur Elon Musk. The car manufacturer was the first to market autopilot cars, but last January the tycoon announced on Twitter that Tesla’s “full autonomous driving ability” will increase within a few months, and predicted that by the end of the year one of its cars will be able to drive alone from Los Angeles to New York without the driver having to touch the steering wheel. Musk has also promised that by 2019 his company will launch a self-driving car of level 5 on the market. Yet the path to this goal does not seem the simplest. For a year now, vehicles pre-equipped with hardware that the company claims to be sufficient for “fully autonomous” driving, are on sale. However, two negative elements cast some shadows on the future of the Tesla-branded driverless car: the company released a six-monthly report showing heavy losses ($ 671 million): Musk said he was “fairly confident” that the current system can achieve “a very similar autonomy to the human level”, but admitted that “if an upgrade is necessary, Tesla will replace the current computers with something more powerful”. In short, not everything seems to go off smoothly.
GOOGLE SELF DRIVING CAR: WAYMO PROJECT
Google has been working on a self-driving car project since 2009. In 2016, it created a spin-off entirely dedicated to the initiative, Waymo, currently led by engineer Sebastian Thrun, director of the Stanford Artificial Intelligence Laboratory and co-inventor of Google Street View. In May 2016, Google signed an agreement with FCA group, according to which the Chrysler Pacifica van, after an initial test phase, will be the first commercial model of self-driving cars branded Google that can be purchased in the United States. The company is filing several patents on techniques for self-driving. “Since the project started in 2009 in Google – we read on the Waymo website – we have collected an amount equal to over 300 years of human driving experience, mostly on city streets. This is added to one billion miles that our drivelerss car drove just in 2016”.
Last September Intel signed a partnership with Waymo. Waymo’s latest vehicles, the Chrysler Pacifica hybrid driverless minivans, already have Intel sensor processing technologies, generic computing and connectivity, enabling real-time decisions for full autonomy in urban situations
As Waymo’s autonomous vehicle technology becomes ever smarter and more efficient, the hardware and software on which it is based will require even more powerful and efficient processing capabilities. Therefore Intel will offer the fleet of Waymo vehicles the advanced processing power required for 4 and 5 autonomy levels.