Navigating uncertainty

Vehicle-to-vehicle and vehicle-to-infrastructure communications have been widely heralded as the next big thing in urban transportation. We take a look at the benefits this could bring

‘If you can’t see my mirrors, I can’t see you!’ So reads the warning sign on the back of so many trucks across Europe and North America, informing following motorists that their own safety relies on their presence being known to the vehicle’s driver.
    
In India they have a local variation ‘Horn Please!’ warning following motorists to make their presence known audibly in the traditional manner. These simple warnings underscore the inescapable fact since the dawn of motorised transportation, that the driver’s line of sight and auditory range have been the primary sensing and monitoring mechanisms for collision avoidance. They also hold a lesson for engineers of future Intelligent Transport Systems (ITS) – if safety related signals are interrupted, the increased risk of a collision is obvious.
    
In recent times automakers have been able to add many more sensing mechanisms to the portfolio of primary safety systems on their products. In the very latest generation of autonomous emergency braking systems, for example, short range radar and camera based ranging technologies such as LiDAR, are used to provide automatic collision avoidance and mitigation benefits, intervening in vehicle emergency braking at a reaction time much faster than would be possible by a human driver.  
    
But even these are seen by many as being merely the tip of the iceberg in terms of what can be achieved by ITS innovations harnessing the availability of satellite positioning, telecoms and WiFi systems. These technologies have had a clearly transformative effect on our lives in their own primary domains of use – in mobile phone communications, satellite navigation and computer networking – but applied collectively in an integrated manner, they offer the prospect of profound changes in transportation.

GLOBAL POSITIONING SYSTEMS
Navigation is perhaps the most obvious of these, where the humble road atlas – once an essential automotive glove-box accessory – is now arguably all but superseded by the satellite navigation system. In the fifteen years since the Clinton administration made the United States’ Global Positioning System (GPS) available on a world-wide basis for civilian applications, our collective reliance upon the signals received from such global navigation satellite systems (GNSS) has extended into almost every aspect of infotainment.
    
While mobile phones have long been able to betray their position via triangulation of signals from network masts, pretty much any smart phone as well as a good number of digital cameras and watches for example, now know exactly where they are at any instant thanks to their in-built GPS receivers. But if we’re happy to use such positioning systems to get an accurate fix on our cars current location and use this to navigate to our desired location, would we be as comfortable in the future to use such external signals as the basis of safety-critical autonomously acting collision avoidance systems?

EXCHANGING SAFETY INFORMATION
Cooperative vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications have been widely heralded as the next big thing in urban and extra-urban transportation. If implemented effectively, traffic management systems and individual vehicles would be able to communicate in order to optimise road network efficiency based on real-time information about each vehicle’s location, bearing, speed and intended destination.     

Personalised routing guidance, safety alerts and speed recommendations to groups of vehicles using instantaneous traffic information would all be possible, and the deployment of technology on vehicles would extend beyond the domain of mere collision avoidance and into the realm of optimisation of traffic movement. After all, if a vehicle knows exactly what traffic is over the brow of a hill, around the next corner, or approaching the same junction and is able to cooperate and communicate with it, the benefits for the safety and efficiency of travel are obvious.

With the consequent increases in effective road network capacity and reductions in localised congestion, traffic would flow more smoothly with fewer stops, thus improving air quality and improving safety. Special priority could be given to certain classes of vehicles, such as emergency services or public transport, and the increased efficiencies of vehicle operation would deliver a potentially substantial contribution to reductions in transport CO2 emissions. Moreover such ITS innovations could generally be implemented at a fraction of the cost of more conventional automotive technologies and traffic management strategies, including the currently politically unfashionable and least environmentally attractive alternative of new road building.

Public acceptability
But even if the communications infrastructure is now becoming available, is society really ready to accept the risks of delegating safety-critical decision making and control to such autonomous systems? Well, the airline industry offers us some fairly promising lessons. At over 10,000 metres altitude and 800 Km/hr, no-one questions that avoiding collisions is an imperative in the operation of a commercial airliner. Modern on-board traffic collision avoidance transponder systems on commercial jets will automatically instruct flight crew to take specified coordinated preventative evasive action if an immediate risk of collision is detected – even if the manoeuvres so demanded run contrary to air traffic control instructions. We accept implicitly that the system is there to protect us from impending disaster and that the instructions of human controllers are potentially fallible and may from time to time need to be over-ridden. In the same way, perhaps public acceptability of autonomous vehicle control systems based on V2V and V2I will ultimately gain acceptance once their value in avoiding or mitigating the effects of road collisions has been demonstrated.
    
But two fundamental considerations remain: how to ensure safety and integrity in all potential modes of failure and what to do in the event that the very channels of communication on which the system is based are interrupted or unavailable?

Fault tolerance
The motor industry is well used to the implementation of development processes for safety-critical control systems and software in which no fail-safe condition can be guaranteed. In such circumstances the philosophy of fault tolerance, as opposed fail-safe operation, is built-in from the concept stage onwards. Typically, systems are classified prior to development in terms of a required Safety Integrity Level (SIL), which defines the nature of the design rules and processes to be followed in development. For example, basic infotainment and informational navigation systems would be at the lowest level – while systems for controlling brake or steer-by-wire that have no mechanical back-up or full authority active driveline control would carry the highest SIL, requiring the most thorough standards of inherent fault-tolerance.
    
In 2009 the development frameworks that apply to safety-related electrical and electronic systems for road vehicles moved a significant step forward with the publication of the draft international standard ISO/DIS 26262 (Road vehicles – Functional safety). This standard aims to bring regulations covering the development of safety critical systems up to date by focusing on the specific needs of road vehicles and encompassing state-of-the-art design processes increasingly used by the automotive industry, including model based control system development.

Integrity of communications
Many processes of safety-critical automotive vehicle systems development have been created with the help of a solid body of knowledge and experience drawn from the aerospace sector, in some cases existing tools such as SCADE that have been directly adapted for automotive application. But ITS-based V2V and V2I system developments, which by definition go beyond the boundary of the vehicle, present an entirely new and more complex development challenge. In the operation of these systems the integrity of data provided by external systems – as well as the mode of exchange between systems – is of crucial importance. And while in-vehicle systems can be considered largely in terms of direct operational considerations, V2V and V2I are inherently more exposed to external threats arising from physical interruption of signals or even malicious attacks.
    
System robustness and resilience in the face of network interruption and signal degradation is thus a completely new development challenge for the automotive industry, taking the concept of fault tolerance to a new and more complex level. While the urban environment offers a potentially rich mix of potential channels of network interoperability – from standard GSM, GPRS and 3G mobile coms to the extensive mesh based WiFi networks controlling infrastructure, such as traffic signals and CCTV monitoring systems – this same environment provides some of the biggest challenges to signal degradation in the form of cross network interference and the obstructions caused by the buildings themselves.  

Testing advances
The need for a dedicated centre at which ITS technologies including V2V and V2I systems can be developed in a highly controlled manner and in complete safety will be addressed by the creation of innovITS ADVANCE, the world’s first such dedicated facility. The ‘city circuit’ of this new facility provides a unique environment for the highly controlled testing and development of ITS innovations. Located in the UK West Midlands, innovITS ADVANCE comprises a comprehensive road network designed to be configurable to replicate an urban driving environment anywhere in the world with numerous junctions, intersections, roundabouts and multi-lane highways. This is equipped with roadside architecture including traffic signals configured for both left and right side traffic circulation, CCTV and catwalk gantries allowing installation of overhead equipment for testing, monitoring and signage.
    
The road network is combined with an open architecture of multi-zoned Wi-Fi, and GSM, GPRS and 3G private stand-alone mobile telecoms networks and WiFi mesh system that can be configured according to the precise needs of particular tests. Each telecom base station transceiver is directional enabling specific enhancement or degradation of regional site coverage.  
    
As the centre’s business development manager Roger Wilson explains: “In creating the innovITS ADVANCE city circuit, we wanted to provide an open-architecture environment in which development teams can recreate exactly the scenarios they require in order to develop new ITS products, services, equipment and systems. It would not be possible – and in many cases neither would it be legally permissible – to carry this type of work out in a real town or city, and it would clearly be impossible to take control of the urban communications infrastructure and carry out all of the ‘what if’ development tests that will be crucial in enabling next generation V2V and V2I innovations to be developed.”
    
But what about the ever-present GNSS signal – how can this be controlled to enable the effects of degradation or denial to be replicated to model the influence, say, of high rise buildings and their associated ‘urban canyons’? Here too, innovITS ADVANCE has some state-of-the-art testing technology available to its customers. “We’ve partnered with NSL to implement their ‘Skyclone’ system,” Wilson continues. “This enables our customers to effectively replicate the effects of a wide variety of urban built environments surrounding our city circuit. So while it has the visible appearance of a network of urban roads surrounded by grassed level ground, the conditions of GNSS denial can be made to precisely replicate those, for example, of a high rise environment similar to New York’s Lower Manhattan or London’s Docklands.”
    
The Skyclone system works by creating a computer model of the track and then sub-dividing each region or neighbourhood in the vicinity of the road network.
    
On the computer screen the model resembles the architectural models used by large-scale property developers the world-over. But in this case, Skyclone uses this 3D model to create a highly resolved spatial map of the GNSS signal attenuation that would be expected of the chosen mix of development across the circuit.  By importing this map to a vehicle based processor between the GPS or Galileo antenna and the on-board receiver, GNSS signals are varied according to the chosen scenario.

Catching up with aerospace
While the aerospace sector might be considered more advanced than road transport in terms of the operation of advanced V2V systems and autonomous operation, it too is struggling with the development needs of such new systems.
    
Only last year there was controversy over the forced withdrawal from service of a regional UK police force’s UAV surveillance drone, which was unlicensed for use in public airspace by the country’s Civil Aviation Authority. The fact that the aerospace industry appears to be struggling to demonstrate the safety case for the telemetry and other systems that support the non-military operation of such remotely controlled UAVs in civilian airspace, is perhaps indicative that the automotive sector is now tackling very similar issues in attempting to implement V2V and V2I systems. In this uncharted territory, the cooperative systems development frameworks used for road transport must be no less rigorous than those that have delivered the safety critical vehicle and aircraft systems that we all enjoy today.
    
“The potential of cooperative V2V and V2I technology to improve the sustainability of transportation through reduced carbon footprint and improved efficiency of road space utilisation is immense,” asserts Wilson. “And when these clear benefits are considered alongside the safety improvements that could be enabled, the case is clearly very compelling. But telecoms companies, automakers and highway authorities need to work together to develop the required systems and prove their safety case in a secure environment where almost all and any real-world scenario can be replicated. With the innovITS ADVANCE facility we have created exactly that type of blank canvass – a web-controlled ‘city’ that can be configured and instrumented in any manner, allowing such new products and innovations to be tested in a plug-and-play manner.”
    
So the trucks of the future – and for that matter the cars, coaches and motorcycles – may not need to carry signs warning following traffic to be seen and heard. If the V2V and V2I innovations envisaged by Wilson and his colleagues at innovITS are developed and brought to market, the vehicle and infrastructure will know exactly where other traffic is, even if the driver is distracted.

FOR MORE INFORMATION
www.innovits.com