The rocky road to safe autonomous vehicles
We often hear from road traffic regulators and the automotive industry that advanced technology, including artificial intelligence (AI), will make our roads safer and more efficient as a result of intelligent infrastructure and fully functioning autonomous, connected cars. There will also be reduced pollution and congestion, no more parking hassles and time for passengers to engage in other activities while their vehicles do all the driving. In theory, this sounds great.
In practice, however, it is far more complex to realize than to theorize, and this utopian scenario is still a long way off. More work still needs to be done to ensure cars are able to recognize objects that cross their path as well as to handle different weather conditions and all the possible outcomes already mentioned.
Recently another accident reported in the US ended in a self-driving car being flipped on its side. While this car was not at fault, it highlights an important issue in the transition to autonomous cars: roads will be filled with a combination of human-driven and driverless vehicles. Not only will humans and algorithms react differently to the same road situation, most people will not accept this technology unless it functions perfectly and is secure from cyber threats.
Several Subcommittees (SCs) of the Joint Technical Committee of IEC and the International Organization for Standardization, ISO/IEC JTC 1: Information technology, are working towards increasing the security, and therefore the safety, of connected cars. Telecommunications and information exchange between systems, comprising near field communications, are covered by ISO/IEC JTC 1/SC 6, IT security techniques, such as lightweight cryptography for vehicles, come under ISO/IEC JTC 1/SC 27, while ISO/IEC JTC 1/SC 38 deals with the cloud computing used in some cars to process, analyze and store the large amounts of data being gathered. Find out more about IEC cyber security work in the article Protecting road vehicles from cyber attacks in this edition of e-tech.
The next major disruptive technology
From healthcare and manufacturing to finance and customer service, many industries are being transformed by artificial intelligence (AI). A market research report by Tractica forecasts that the annual global revenue for AI products and services will be worth USD 36,8 billion by 2025.
Behind AI is a relatively new computing model called deep learning. Powerful graphics processing units (GPUs) allow machines to learn, by gathering and analyzing huge amounts of data, in order to enable them to make autonomous decisions.
As this technology advances, numerous smart sensors, embedded connectivity applications and location-based big apps related to the running of the car and to its surroundings (other vehicles, pedestrians, traffic and weather updates), will gather data. These will then be exchanged and used throughout the intelligent road infrastructure to improve road safety and efficiency.
Standardization every step of the way
A number of IEC technical committees (TCs) and their SCs produce International Standards which contribute towards ensuring the safety, reliability and interoperability of car components. Some of these include lamps and related equipment (IEC TC 34), sensors (IEC TC 47), EV charging (IEC TC 69), audio, video and multimedia systems and equipment, such as dashboard touchscreens (IEC TC 110). Additionally, IEC International Standards for cyber security will be applicable to the software and hardware in connected vehicles.
AI could enhance road safety and efficiency
Deep learning, combined with biometrics, is already changing the driving experience. Some smart cars know the drivers before they get in, thanks to pre-programmed facial recognition apps, while others require fingerprint authentication in addition to a key to start the motor, adding a layer of security. For more information about biometrics and connected cars, read Biometrics for consumer markets in this edition of e-tech.
Many new cars already contain AI technology, which can enhance safety and security; for example, driver assistance systems, automatic braking, smart cruise controls, pedestrian and cross-traffic alerts as well as collision avoidance systems. Facial recognition and external car sensors track external objects or pedestrians, while voice recognition alerts drivers to the presence of the objects that have been detected.
The automotive industry is working towards fully connected cars being able to talk to one another using vehicle to vehicle (V2V) technology, and to intelligent traffic systems through vehicle to infrastructure (V2I) technology. As a result of this interaction, cars will adjust speeds and general driving automatically so as to suit the surroundings. Increased communication, access to real-time traffic and weather conditions and knowing what drivers in the immediate vicinity and further away are doing, will – hopefully – lead to safer, more efficient roads.
Research by Gartner suggests that by 2020, there will be 250 million connected vehicles on the road. This connectivity could open the door to an endless list of in-vehicle services. For example, they may include intelligent car assistants who can find the nearest parking (read more in article Smarter than the average lamppost in this issue of e-tech) or advise passengers of the closest restaurants, petrol or charging stations or shopping facilities. Apps which offer real-time engine monitoring will be able to flag up potential issues, send a message to the nearest garage and arrange repairs as rapidly as possible, thereby avoiding and minimizing the effects of mechanical breakdown.
Nothing works without connectivity
Even if all of these apps and services are developed, using Wi-Fi, LAN or the cloud, they will only work if they are connected to the Internet of Things (IoT). The IoT will allow the requisite vital exchange of information between apps within the car and external service providers and infrastructure networks.
The work of ISO/IEC JTC 1: Information technology, covers “the specification, design and development of systems and tools dealing with the capture, representation, processing, security, transfer, interchange, presentation, management, organization, storage and retrieval of information”.
A growing number of consumer devices and systems are becoming part of the IoT. In light of this, ISO/IEC JTC 1/SC 41 has recently been established to develop International Standards for IoT and related technologies.
Discussing the future networked car
During the 2017 Geneva Motor Show, a workshop for the Future Networked Car was organized by the International Telecommunication Union (ITU) and the United Nations Economic Commission for Europe (UNECE). Regulators, standardization and certification organizations as well as automotive industry, technology and cyber security experts gathered to discuss the changes that AI and machine learning will bring to vehicles and transport, different aspects of connected vehicles and automated driving and how to mitigate cyber security threats.
Automotive industry: back to the drawing board
As well as affecting how cars run, AI technology is contributing towards changing the driving experience radically and forcing the automotive industry to rethink its business model for personal transport. This evolution was highlighted during the workshop.
Against the backdrop of increases in aging populations, car running costs, pollution, congestion and parking inconveniences, there is a very strong case for reassessing personal transport, particularly in urban areas.
“We are considering new business models that would merge individual mobility and public transportation”, said Anders Eugensson, Director, Governmental Affairs, Volvo Car Group, Sweden.
Eugensson went on to explain the possibility of running fleets of cars for ride-sharing services, allowing people to use a vehicle without owning it. Another option would be to offer peer-to-peer sharing: participating car owners rent out their cars when they are not using them. Renters access nearby and affordable vehicles and only pay for the time they use them.
Increasingly, cars are driven by technology, and car manufacturers are considering new business services such as ride sharing instead of focusing solely on selling cars. As a result, in the last couple of years, many partnerships have been formed between automobile makers and tech companies. Data gathered by CB Insights shows that 2016 was a record-breaking year for partnerships between tech companies and car manufacturers, amounting to a total of USD 450 million invested across 33 corporate groups.