Connected vehicles: wireless control for safer stops and crash-free travel
Image credit: Alamy
Wireless control techniques will cut accidents on the road and in the air - but can connected vehicles be safe from hackers?
For police trying to stop offenders in stolen cars the weapon of choice has become the ‘stinger’ - a simple chain of spikes designed to blow out the tyres of the oncoming vehicles. But they take time to deploy and lead to risks for the police stretching them across the road. In the attempt to avoid a stinger spread across half of the road in a high-speed chase, a teenage vehicle thief mowed down Merseyside police constable Dave Phillips in October 2015. The European Union-funded SAVELEC project was set up to find a safer way of bringing offenders to a halt. The researchers borrowed an idea from electronic warfare that uses a sudden electromagnetic pulse (EMP) to disable the circuitry in the car that controls the engine.
“The aim is to create a disturbance that causes a temporal malfunction in the logic operation of the vehicle,” says Dr Marta Martínez-Vázquez, SAVELEC co-ordinator and senior researcher at IMST, an industrial research, engineering and design company based in Germany. “The most common situations are escapes from police controls, with too much speed leading eventually to loss of control of a car,” Martínez-Vázquez adds. She says based on data from the past three years in Saxony-Anhalt in Germany, “we identified about two cases per week where an EMP safe-stop of a car would have saved at least the health or even one or more lives”.
The effect of the EMP on the vehicle is akin to someone pointing an electromagnetic pulse close to your PC and forcing the operating system to a halt or to bring up the “blue screen of death”.
UK-based electronics company e2v is developing an electromagnetic pulse generator called RF Safe-Stop, designed to be used not just against road vehicles at ranges of up to 50m but also boats and unmanned aerial vehicles (UAVs).
“We are currently working in partnership with a customer to counter UAVs. UAVs, for example, can be stopped at up to 400 metres,” says Andy Wood, product manager at e2v.
A vehicle-stopping EMP cannot be as indiscriminate as one used in electronic warfare, where the sole intention is to disable everything in range. In these devices, the pulse’s energy is generally confined to specific parts of the RF or microwave spectrum. The RF Safe-Stop, for instance, can generate directed beams of RF in the S and L bands, which reach from 1GHz up to 4GHz.
“It is important to maintain a final level of control via human intervention,” Wood says.
Researchers in the SAVELEC project designed their system to avoid permanent damage to vehicles using their high-power microwave generators. The researchers claim the microwave pulse will not set off airbags or ignite fuel accidentally, for example. They used simulations to test driver reaction to the remote-controlled stop.
“None of the drivers in the simulator overreacted or endangered themselves, or others, with critical manoeuvres. Almost all drivers steered the car to the hard shoulder after they realised that the engine had stopped,” claims Martínez-Vázquez.
The project partners then designed and assembled a prototype car-stopping device, which they demonstrated on an open field track during the project’s final review. The system exceeded expectations, halting a moving car as well as stopping the engine of a stationary one.
As vehicles become more autonomous, the authorities will have more subtle means of keeping them under control than trying to confuse their circuitry. Vehicle-to-vehicle (V2V) communications will make it possible for cars to detect hazards long before the driver does. Automated-assisted travel is almost here whether you’re ready or not.
“V2V enables cars to talk with other cars by sending important safety messages such as location and speed between approaching vehicles, and then alerting the driver to help avoid or mitigate a crash,” says Alan Adler, manager of autonomous vehicles and R&D communications at General Motors (GM). For example, signals from a car hidden by a bend can tell the driver or the on-board systems not to attempt to overtake.
“Humans are responsible for about 90 per cent of crashes,” Adler claims. He cites statistics from the US National Highway Traffic Safety Administration, which estimates that if all vehicles were equipped with V2V it would prevent 80 per cent of crashes caused by drivers who were not impaired by alcohol or by fiddling with gadgets. However, the collision of a Tesla Model S with a trailer indicates that autonomous vehicles will not necessarily be accident-free.
The car is not alone in getting remote-control assistance. Two major next-generation air traffic management systems are being developed to enable software to help traffic controllers on the ground when planes approach the runway in heavy traffic conditions - a growing problem as airports try to service more aircraft with the same number of runways.
With these systems there will be fewer manual operations and less communication needed between controllers and pilots. Procedures known as performance-based navigation (PBN) will allow planes to take more direct flight paths, which will help reduce overall fuel consumption. Flight paths assisted by automatic control will keep aircraft at their most efficient altitude for as long as possible before starting a smooth approach to the airport.In Europe, the Single European Sky Air-Traffic Management Research project (SESAR) is being led by EuroControl and other partners, while the Next-Generation Air Transportation System (NextGen), the US equivalent, is being driven by the Federal Aviation Administration (FAA).
As part of this system, the Terminal Sequencing and Spacing (TSAS) tool has undergone testing by Nasa and the FAA. In April 2015, FAA air-traffic controllers and traffic managers performed simulated busy-arrival operations into Phoenix Sky Harbor International Airport. The tests simulated real-world events, such as missed approaches, emergency landings by other aircraft and runway-configuration changes.
“Without TSAS improvements, only 42 per cent of the arrivals were able to conduct fuel-efficient descents from cruise to landing. In other words, 58 per cent of the arrivals were interrupted by the air traffic controller in order to merge and space with other aircraft,” says John Robinson, chief engineer of the Nasa group that developed TSAS. From 2018, the system will start being deployed at a number of busy international-grade airports in the US, including Atlanta, Denver, Houston, Las Vegas, Los Angeles, Phoenix and San Francisco.
Handing over control
As commercial aircraft systems develop they may acquire even more systems intended to prevent them diving into the ground or colliding with each other. One example is the Automatic Ground-Collision Avoidance System (Auto-GCAS) developed by the US Air Force Research Laboratory (AFRL), Nasa’s Armstrong Flight Research Center, the Air Force Test Center (AFTC) and Lockheed Martin’s Skunk Works.
Incidentally, Lockheed Martin’s Skunk Works previously developed the most advanced aircraft in aviation history: namely the U-2 plane, the Lockheed SR-71 Blackbird and the F-117 Nighthawk stealth aircraft.
Although military aircraft have for some years incorporated voice warnings to tell pilots when they are executing a dangerous manoeuvre, Auto-GCAS takes direct control. The system kicks in at the last moments before a potential crash, performs a life-saving manoeuvre and then returns control to the pilot. The system could help reduce the number of deaths arising from aircraft accidents involving ground contact, which result in around 100 deaths each year in the US alone. Auto-GCAS has been installed on more than 500 US Air Force F-16s worldwide. The system was credited with saving the life of a pilot in November 2014 and two more since.
Saving money... and lives
The system predicts the aircraft’s trajectory over an on-board digital terrain database. It applies the following rules: don’t make the situation worse; activate when an aware pilot would normally respond; manoeuvre the aircraft as a pilot normally would. However, as military pilots often have to execute risky manoeuvres, it does not take control every time it predicts a collision.
“The system has a function that allows the pilot to override it at any time,” says Lockheed Martin programme manager Ed Griffin. He adds: “In addition to preserving precious life and valuable military aircraft, the integration of Auto-GCAS will result in significant savings of over $4bn over the next 25 years according to the Office of Secretary of Defense.”
An Auto-GCAS version 2.0 is now on the way. Nasa is working to expand Auto-GCAS to lower-performance general aviation aircraft, including UAVs. To test the software, the team used off-the-shelf consumer electronics to take advantage of their fast, low-power processors. A smartphone on the UAV stored the terrain data needed to navigate safely.
“The Air Force is also developing a related system to automatically prevent air collisions: Auto-ACAS, which will ultimately be combined with Auto-GCAS to provide a fully integrated collision-avoidance system,” says Griffin.
A system like this can, however, still be deactivated or over-ridden. It won’t stop someone from intentionally hitting the ground, such as in situations similar to the ill-fated Germanwings Flight 9525. In March 2015 the co-pilot locked himself in the cockpit and flew the aircraft from cruising altitude into the side of a mountain in the French Alps.
Fully autonomous control over cars seems more likely in the near term, with multiple projects combining advanced computing with V2V communications. In the Netherlands, an autonomous shuttle bus, the WePod, is now in regular service along a 200-metre stretch of public road in the town of Wageningen, and in Perth, Australia, a French-made driverless electric bus was trialled this April on private roads. In the UK, tests are taking place using driverless pods. The global market for automated vehicles is expected to reach 44 million by 2030.
Elon Musk, the founder of Tesla, says his electric cars will be completely driverless in two years. Other self-driving proponents have touted a vision that seems better for the environment and us by taking human error out of the process.
As well as safety, fuel usage provides a push for vehicle autonomy. John Dolan, principal systems scientist at Carnegie Mellon University’s Robotics Institute, says: “Autonomous cars can provide more gradual and controlled acceleration and deceleration, thereby reducing fuel usage.”
The US and China are working on more intelligent traffic lights that are intended to cut congestion and fuel consumption as well as driving times. The idea is that V2V communication can inform vehicles approaching a junction of the ideal speed to be sure of a green light - and potentially even let cars cross each other’s paths at a safe distance. Instead of braking to a halt, much like an aircraft on a remote-guided flight path, it would slow to an appropriate speed and glide across. Non-autonomous vehicles would need to wait until the light in their lane turns green.
But such interconnected cars come with their own risks. What if the threat to traffic stops being the drivers but bystanders armed with electronic weapons?
This type of remote infiltration has already been tested by a group of US-based researchers in Virginia last year. They found it relatively easy to remotely hack into a driverless car’s control system, identifying several vulnerabilities in cars with wireless technology.
“The surest way to avoid hacking into self-driving cars is to operate without using wireless communications. Much autonomous driving performed to date does just that,” says Dolan.
V2V is almost certain to be needed to make the most of autonomous and semi-autonomous driving. Securing all communication vulnerabilities to hackers is and will continue to be an uphill struggle. “Computer security is a cops-and-robbers affair, which never ends, with continual attempts by each side to outsmart the other,” Dolan adds.
Car companies have started to work together to deal with the potential hacking challenge. GM, among others, has joined Auto ISAC: an industry-wide effort to identify emerging threats and potential adversaries.
Tom Wilkinson, who is responsible for safety and cybersecurity communications at GM, says: “We also recently announced the GM Security Vulnerability Disclosure Programme, through which security researchers can inform GM of bugs or vulnerabilities via a security website portal hosted by HackerOne. We don’t discuss details of security measures for a simple reason - we don’t want to give potential adversaries useful information.”
The question will be whether they also build in anti-EMP defences. If more subtle hacks don’t work, a blast of high-energy microwaves may be just the thing to bring automated vehicles to an unexpected halt. *