РУ EN

Hot line phone +7 495 660 75 89

logo

Cars' Calculating Mind: How Artificial Intelligence Is Changing Mechanical Engineering

30 January 2018

Artificial intelligence developers and automobile concerns are forming new alliances. At their current level, computer vision technologies and the latest software will allow for mass production of self-driving vehicles in the near future.

What is the current market situation? When can we expect to see safe autonomous cars on the road? INNOPROM VIEW answers these questions.

Intelligence races

Artificial intelligence (AI) technologies are changing the structure of industry. While early 2010s saw the first announcements and tests of computer-based intelligence, current developments, such as Advanced Driver Assistance Systems (ADAS), are prompting the launch of AI into mass production and harbinger major changes in mechanical engineering.

At the end of 2017, the International Robotics Center (IRC) compiled a ranking of key companies working in artificial intelligence development. The list contained 11 organizations. Israeli Mobileye claimed the top spot. This company is the current leader in the computer vision market. For example, Israeli specialists developed Mobileye hardware and software complex, which warns drivers about potentially dangerous situations on the road and ensures safety driving distance control. Mobileye utilizes unique algorithms to recognize and analyze dynamic road environment while a powerful processor continuously processes collected data and identifies accident-prone situations. Intel corporation acquired the Israeli company in 2017 for $15bn.

Cognitive Technologies, a Russian company, also partakes in the artificial intelligence race. It holds the second place in the IRC ranking of global AI developers.









"The majority of car manufacturers are currently using Mobileye products. Still, over the past several years, Cognitive Technologies, the only Russian company on the list, has been advancing fast. Over a short period of time, the company managed to create a competitive intelligent product which is superior to its counterparts in a number of areas, including its ability to perform under adverse weather conditions and on bad roads," ― IRC points out. 









The Russian developer created Cognitive Smart Situation Capture technology, which allows for precise forecasting of the situation on the road and produces safe driving scenarios taking into accountto various circumstances, including critical conditions.











"Neuron net tools can be also applied to situation forecasting. If a prognosis matches reality, relevant data can be transferred over to the neuron net which will use accurate forecasts for self-training during autorobot movement,"
― Yury Minkin, Head of Autonomous Vehicle Development Department at Cognitive Technologies, shared.









More than 10 Western and Asian brands have already expressed their interest in products developed by the Russian company. Cognitive Technologies was recognized as Company of the Year at the Consumer Electronics Show CES-2018 in Las Vegas, NV. 

Experts predict that ADAS sales will spike in the near future. According to Lux Research analysts, in 2020 market volume will amount to $10bn, while corresponding Yano Research Institute estimates are as high as $75bn. 

Corporations Welcome Startups

Major large corporations, such as Volkswagen, Daimler, Tesla, BMW Group, Toyota, etc. have announced their interest in autonomous vehicles. Top mechanical engineering players are ready to spring to action as they anticipate a sharp increase in driverless car production and sales. In order to launch mass production of autonomous vehicles, large corporations invest in AI development and create their own systems. 

For example, General Motors acquired Cruise Automation startup for over $1bn. On top of that, the automobile concern was the first to announce the launch of large-scale sales of driverless cars: in 2019, the company will present its Cruise AV e-car with no steering wheel and no pedals. 









"When you don't have a steering wheel, it makes no sense to talk about an airbag in a steering wheel. So what we do instead is put in an airbag that mirrors the right front passenger side, and show it provides equivalent safety,"
— Paul Hemmersbaugh, GM's policy director for autonomous vehicles, stated.












The car is equipped with two synchronized computer systems. According to GM data, if one system fails, the other assumes full control of the vehicle. "Within the computers' operations, Planning determines the desired vehicle behavior. It accounts for road rules and plans routes for the car to travel from trip origin to destination. It chooses routes to optimize efficiency and safety and to route the car only on streets within its capabilities. If something unexpected happens, Planning has multiple backup plans," General Motors informs.

Ford is also growing its business through technological startups. The corporation develops "the brain" of autonomous vehicles, i.e. their virtual control system based on machine learning technology. In order to tackle this goal, the company contracted specialists from Argo AI, a young company working in robotics and artificial intelligence. Argo AI was created in 2016 by former Google and Uber employees. In 2017, Ford purchased the majority stake in the company. Argo AI developers will cooperate with Ford staff to create a new software platform for driverless vehicles which is integral to launching Ford's new autonomous car by 2021. Between now and 2021, $1bn will be invested in Argo AI.









"We are at an inflection point in using artificial intelligence in a wide range of applications, and the successful deployment of self-driving cars will fundamentally change how people and goods move,"
— Bryan Salesky, CEO of Argo AI, said.









Let us not forget that Ford tested its first autonomous car back in 2013. The pilot model was equipped with a car computer, and it ran on special software developed at the University of Michigan. In 2016, the concern announced an upgrade: improved sensor performance and software platform capacity thanks to switching to new hardware. In the next several years, the company intends to finetune and complete the final version of the autonomous vehicle. 

In purchasing AI products, car manufacturers seek to accumulate the resources required for mass production and sales of driverless vehicles. According to forecasts, almost 2mln autonomous models will be sold in 2020, 12.3mln in 2025, and over 30mln in 2035. 

So far, the Russian industry has made its first small steps in this direction. In 2015, Russian truck manufacturer KamAZ carried out its first tests of driverless vehicles at its pilot site, and in 2017 the company presented its project developed in cooperation with the Central Scientific Research Automobile and Automotive Engine Institute "NAMI." KamAZ and NAMIdemonstrated their driverless Shuttle bus which can independently transport passengers and accelerate up to 40 km/h. The Russian-made novel product was presented before the President of the Russian Federation Vladimir Putin and the Minister of Industry and Trade Denis Manturov. 
At INNOPROM-2017, the President of the Russian Federation Vladimir Putin explored the capabilities of the Russian-made driverless Shuttle bus / Photo courtesy of kremlin.ru

From Connection to Autonomy

Asia and the West are seeing a general increase in the popularity of the so-called "connected vehicle" segment. Connected models can be perceived as an intermediary stage between technology controlled exclusively by humans and robotic vehicles. Connected cars exchange data with other vehicles, networks, and services. Such objects are a part of the Internet of Things ecosystem while maintaining human control.

According to estimates provided by Statista, a German market research firm, the largest markets of connected vehicles are the U.S. ($8.2bn in 2017), China ($3bn), Germany ($2.6bn), and Japan ($1.7bn). The size of the South Korean market is gauged at $250mln. 

Json & Partners analysts report that as at 2017 there are 330,000 driverless vehicles worldwide. 

In order to monitor the development of this market, the US-based Bloomberg Philanthropies Foundation and the Aspen Institute created an interactive map of cities which are currently testing or intend to test autonomous cars. As this material was prepared for publication, more than 70 cities worldwide were testing driverless technologies or announced upcoming tests. Another 30 municipalities stated that they were making arrangements for driverless transport checks.

In Russia, Yandex is implementing its autonomous project. The company is exploring the capabilities of its navigation system: in the end of 2017, Yandex tested a driverless car in the wintertime in the Greater Moscow area. At the moment, it is unclear when the technology is going to become a mass product. According to the Ministry of Transport of the Russian Federation and the Federal State Unitary Enterprise "Zashchitainfotrans,"in seven years Russia may develop a unified system designed to monitor all driverless vehicles. 

Denis Manturov, the Minister of Industry and Trade of the Russian Federation, believes that given the current state of affairs and significant risks, autonomous cars need to be tested before they are allowed on public roads. 












"Implementation will proceed gradually. Partially automated vehicles will not require major changes in road infrastructure, but the introduction of fully driverless cars will definitely necessitate the installation of additional sensors and electronic elements,"
— Denis Manturov, Minister of Industry and Trade, stated in his interview with RIA Novosti.












South Korea (INNOPROM-2018 Partner Country) is building an entire city for autonomous vehicle tests. 2017 saw the official opening of K-City pilot site. The federal budget will invest $10mln into the facility. The site spans over 32,000 sq.m. and allows for modeling various driving situations, e.g. tunnels, intersections, parking, construction sites, dirt roads, etc. 

The project was launched in the spring of 2017 and is expected to be completed in the first half of 2018. Private companies, researchers, and industrial startups will have access to the pilot site. 
 
 
"I hope the world-class facility K-City becomes a stepping stone to our country's autonomous driving technology,"
— Kim Hyun-mee, Minister of Land, Infrastructure and Transport of South Korea, pointed out.












Safety Procedures

Automobile concerns are currently emphasizing the safety of artificial intelligence for driverless vehicle owners. First autonomous cars have already made road accident reports. In 2016, a Google car caused an accident for the first time: it hit a city bus in California when trying to go around sand sacks left in the line of traffic. The collision damaged the left side and front wheel of the bus; no passengers were hurt. The same year marked the first fatal accident when a Tesla set into autopilot mode got into a traffic accident. 

Representatives of global concerns meet at INNOPROM strategic sessions to share their visions of AI development and the future of mechanical engineering. In 2017, participants of the Artificial Intelligence and Industrial Machine Training Technologies session explored opportunities offered by smart cars and related issues. Japanese Mitsubishi Electriccorporation demonstrated its smart technologies, and KamAZ presented the first Russian driverless bus. We invite you to speak at or attend our discussions held as part of INNOPROM-2018 Business Program. 
All news
  • Sponsors
  • Media partners
  • Organizers

  • Operator
  • TECHNICAL SERVICE
  • Recommended builder
  • Service Partner