Improvements in safety and network capacity are two of the principal benefits expected from the emergence of driverless road vehicles. The use of sensors, communications technology and processors will reduce the number of collisions and enable more vehicles to use existing road space.
These expectations mirror the safety and capacity benefits that airlines and railways have achieved through technological improvements in control systems. Network Rail chief executive Mark Carne now proposes to accelerate the ‘digital railway’ revolution. Central to this will be exploiting in-cab signalling – where the train is in near-continuous contact with the signalling centre – to enable trains to follow each other more closely with no adverse impact on safety.
However, there is one obvious difference between the parallel paths for road and rail technology. The new rail signalling replaces an already regimented signalling system where even minor incursions into the allocated safe separation zone are documented and investigated. The road equivalent, on the other hand, will replace a system where incursion into the prescribed safe separation zone is the norm.
On motorways, for example, the separation specified in the Highway Code is 96m at 70mph, but vehicles commonly follow each other at that speed with less than a quarter of that separation. Similarly, joining a busy motorway or moving from the inside lane to the middle lane would often be impossible without reducing the space between two other vehicles to less than the specified safe separation.
Autonomous vehicles should be able to follow each other with considerably less than the current Highway Code minimum safe separation. Modelling suggests that closer spacing will produce big increases in road capacity, but will that spacing be closer than the theoretical or the observed spacing on motorways today?
David Davies, executive director of the Parliamentary Advisory Council for Transport Safety, says the subject was raised at a recent event on driverless road vehicles. “They were quoting 25% improvements in capacity, certainly on motorways, because of the closer following and the more constant speeds – it’s also about the stop-start nature of the traffic.”
Alex Burrows, Atkins’ technical director for highways and transportation, recalls capacity improvements on the M42 when a ‘managed motorway’ scheme regulated traffic speeds. He says: “The size of the variance between the slowest and fastest is reduced. If more vehicles are travelling at the same speed, capacity is increased.”
Similarly, London Underground’s high track capacity stems from each train following the same speed profile.
Daniel Jonas, Atkins’ head of innovation for rail, cautions: “What will determine capacity are the bottlenecks. In aviation, this is the runway. On railways it’s often the station. It’s all very well saying you can have lots of trains in the white space but dwell time at the station might become the bottleneck.”
The obvious parallels for motorways are interchanges, where true automation will require vehicles to negotiate roundabouts and safely join the motorway with no human intervention. If the mandated separation for those manoeuvres is greater than the separation that drivers currently allow, there is potential for queues on slip roads to affect traffic flow on motorways and roundabouts.
Lee Woodcock, director of intelligent mobility at Atkins, says: “The autonomous decision-making system could make it possible to develop all sorts of interchanges. It may well be that roundabouts would have to change.”
Burrows adds: “We may not need roundabouts. Audi is testing a system where cars are looking three or four traffic lights ahead and are optimising the journey so that everyone can get through at a regulated pace, to optimise journey time.”
The rail industry has decades of experience of driverless trains, but on segregated rights of way where the vehicles have similar characteristics and a small number of routeing permutations. This year the UK is due to start trialling ‘platooning’ of lorries – vehicles with similar physical and journey characteristics.
However, even lorry platoons on motorways will not have an exclusive right of way. The trials will have to address interaction between a convoy of lorries autonomously following their leader and the human drivers of other vehicles. For example, how will the platoon react when a motorist cuts in between lorries in preparation for the next motorway exit?
Davies believes this interaction between humans and machines is a key challenge for developers of autonomous vehicles. “If you have a convoy of vehicles travelling 50 miles up the motorway, joining and leaving the convoy is going to be more tricky.
“The technology is fantastic and it’s going to get cleverer and cleverer. What do we want it to do? Are we going to leave that to the market, and how does it work with people?” He points out that pedestrians and cyclists use many roads and therefore autonomous vehicles will always have to account for human behaviour, even if we reach a point where all motorised traffic is autonomous.
Woodcock sees autonomous vehicles evolving hand-in-glove with other technologies, such as now-common satellite navigation systems. These could deliver safety messages to drivers, such as alerting them to a platoon of lorries and advising them not to start overtaking because the exit they require is near.
Jonas cites the Toyota Prius dashboard, indicating current fuel consumption among other things, as a sign of things to come when the interface between machine and driver could be ‘gameified’ to provide realtime information and guidance. Burrows agrees, pointing to a Honda dashboard that allows drivers, through careful driving, to grow a ‘virtual forest’ as their miles per gallon improve.
Again, railways are a step or two ahead. Many trains now have driver advisory systems to optimise fuel consumption. The digital screen of the new in-cab signalling provides information about the route ahead and advised speed. In the event of overspeeding, a red zone appears on the speedometer and the computer will take control of braking unless the driver quickly applies the brakes.
Railway signalling is rooted in a highly regulated technical environment, which to some extent will need to be mirrored for autonomous road vehicles. In-cab railway signalling stems from the European Union’s desire to harmonise system standards and specifications across member states, where signalling is idiosyncratic.
In comparison, the road system already has a greater degree of international standardisation. The Vienna Convention on Road Signs and Signals created a framework in 1968, and it has long been in vehicle manufacturers’ interests to ensure that their designs meet safety requirements in many countries.
Evolution of driverless road vehicles will require new, additional standards. It also increases the duty of care on the providers of road systems, because the person responsible for each vehicle will become a passive consumer or passenger. Experience from the air and rail industries suggests a requirement for centralised authorities, similar to the UN’s International Civil Aviation Organisation or the UK’s Rail Safety and Standards Board, to define and update standards for autonomous road vehicles and the infrastructure they use.
Davies does not foresee the UK requiring a new body for such tasks. “There are already standards set for vehicles, mostly at a European level for us. I don’t think it necessarily needs a whole new administration.”
Woodcock concurs. “The issue about standards is quite interesting,” he says. “A lot of the vehicle manufacturers are well progressed in terms of autonomous capabilities and a lot of work has to be done around standards. We could risk tying ourselves in knots on safety rather than trusting vehicle manufacturers to do what they’re meant to be doing.
“There are working groups in the EU around standards. We are going to need some. We could learn from other sectors.”
Other modes of transport are not the only places offering useful pointers, according to Jonas, who says: “We will need to start learning from what happens in the defence sector, where you have many systems out in the field from different manufacturers which have to interact. We will need a system of systems.”
Driver-machine interface in a mainline train: similar equipment could be developed for road vehicles
Intriguingly, insurance companies could play a key role in the emergence of standards. The communications technology that makes autonomous vehicles practicable could also feed data to insurers, theoretically enabling each driver’s premium to be calculated according to the risks he or she takes.
Established drivers will need some re-education to adapt to the growing presence of autonomous vehicles, even if they intend never to own such a vehicle themselves. Will they need formal training, similar to the conversion courses for train drivers to use in-cab signalling?
Woodcock believes the son of satnav could obviate this. “Already people’s cars advise them where to go. There’s a further shift in behaviour required for you to completely trust your car. You could see an example where the standards might be altered for safe driving in a semi-automated environment. The real incentive here will probably come from the insurance companies.”
The above hypothesis of a car approaching a lorry platoon shortly before its required motorway exit provides an illustration. If the motorist ignores a warning from the car’s advisory system of insufficient time to overtake the entire platoon, this could place a black mark on the driver’s insurance record.
The data flow from vehicles also unlocks the door to road pricing. The subject has proved too contentious for successive transport ministers, even on a revenue-neutral basis. The need for such a system could become more pressing if, as many have suggested, driverless technology inures motorists to delays because they can watch a movie or chat on the phone or social media while their car crawls through congestion. Even if money can be found for additional motorway capacity, there would be major implications for local roads and streets.
Commercial airlines have decades of experience of yield management, charging premium fares for popular flights and filling spare capacity at bargain prices. Bus and train operators are well versed in offering lower fares for off-peak services than during the commuter peaks. The spread of smart ticketing and m-ticketing, where ticket details are displayed on the passenger’s mobile phone, will theoretically enable operators to charge a different fare for every train during the day between two given stations.
Burrows predicts that the flexibility of smart tickets, exemplified by Oyster validity on all forms of public transport in London, could broaden to embrace mobility in general, as cars come into the fold of communications technology. He says that many young people, in particular, are uninterested in owning cars, but would value being able to use a car when that would be the most suitable mode of transport for a particular journey.
Jonas agrees: “People will have the opportunity to purchase mobility in a very similar way to the way we buy a mobile phone contract,” he says. “The customer will purchase the means to go from A to B without owning a car or a season ticket for a particular train between two stations.
“Their mobility needs would be bought more like a broadband bundle – a package including maybe 20 commutes per month, eight hours of car club and bike hire.”
Burrows says: “It helps make the transition to this alternative business model around car sharing and car hire and hopping from one mode to another. The art is to have a more dynamic service available to incentivise people to make decisions. Steering them towards the options helps not just to give them a satisfactory journey but also to manage the network more efficiently. This is demand management on a sophisticated level.”
Woodcock expects autonomous vehicles and the accompanying technology to open new avenues. “There’s a risk that it will attract quite a lot of demand onto the network. I would rather see the opportunity being taken to give streets back to the public and improve the public realm.
“With this new technology, the Highway Code will need to adapt to this fundamental change to the system.” Together with complementary changes to infrastructure, this creates opportunities to make highways safer and improve journey times, he says. “Nothing should be ruled out. It would be a shame to miss those opportunities because of sticking to previous experience.”