The potential opportunities – and serious challenges – of using satellite positioning systems on the railways came under scrutiny this week at a conference designed to push adoption of the technology up the rail industry’s agenda.
While already in limited use, the industry hopes global navigation satellite systems (GNSS) such as GPS and Europe’s Galileo could play a big part in future safety-critical applications such as signalling and collision avoidance.
But it became clear that the railways present unique technical challenges for GNSS not experienced in other transport sectors such as sea or air, and is likely to need ‘augmentation’ by othersystems where safety is at stake.
The London event, organised by Pinpoint Faraday — set up to help develop GNSS in the UK — and the Rail Safety & Standards Board, was the first major forum to bring together rail industry heavyweights and the wider satellite technology community.
The results of two trial projects designed to test GNSS in the rail environment were outlined for the first time, offering a mixed picture of its effectiveness.
One was a 25-week trial of the accuracy of GPS operating on a high-speed train running between London and the west country and South Wales. It involved space software specialist SciSys and rail engineering group Parsons Brinckerhoff, with the support of the British National Space Centre.
Gareth Close of SciSys said the satellite system proved a success in delivering sufficiently reliable position and time fixes across the trial area.
To his surprise, Close said it also worked deep into centralLondon. ‘We had valid positioning all the way to the buffers at Paddington,’ he said. ‘There are many tall buildings, so it’s a poor environment for GPS — but it worked.’
The system also seemed to function well in the presence of potentially disruptive rail systems such as overhead power lines and the electrified third rail.
There was, however, a problem with ‘cold starts’ when the train left Paddington after standing at the station, and failed to secure a valid GPS signal until it reached the west London suburbs.
Close suggested these and other technical issues could be resolved through augmentation with alternative systems.
The Paddington trial did not, however, attempt to overcome the most obvious drawback to widespread use of GNSS on the rail network — what happens when a train is in a tunnel.
The conference was told that accounting for time spent underground — possibly with a surrogate GPS signal source known as a pseudolite, or by switching to an alternative position sensing technology — is a key question for the industry.
Simon Atkinson, a location and tracking specialist for R&D company Roke Manor Research, claimed that predicting the performance of GNSS across the UK’s vast rail network presents significant challenges. Alongside poor reception areas such as tunnels and cuttings, there is also the problem of multipath, the distorting effect of signals bouncing off buildings and other rail traffic.
In the second trial Atkinson, tested a standard GPS unit on a train running between Southampton and Winchester, and then in a car on the equivalent stretch of the M3, which runs virtually parallel with the rail line.
While the car-based system secured a consistently good signal, the rail journey saw wide variations, with significant dips through cuttings and the complete loss of signal through tunnels.
‘The need for augmentation systems becomes clear from these types of results,’ said Atkinson.
Roke Manor has developed an electromagnetic modelling system that can predict the behaviour of GNSS signals, including the effect of multipath, on the rail network.
‘If GNSS is going to be used for safety-critical applications we have to be able to predict its performance,’ said Atkinson.
Sing from the same timetable
The notoriously fragmented rail industry was warned it must sharpen up its act if it wants to fully exploit the potential of GNSS.
Prof Paul Cross of University College London told the satellite positioning conference that the industry needs a united front if it is to take full advantage of the opportunities the technology will bring in the future.
Cross said the launch of Galileo, upgrades to GPS and the appearance of other satellite navigation systems could greatly improve the performance of GNSS by 2010.
‘We can expect at least 60 and possibly 70 or 80 satellites, and the appearance of quite a significant supporting infrastructure,’ said Cross.
This would lead to increased signal availability and reliability, greater accuracy and less interference.
‘The railway might want to set up its own GNSS structure, but it’s not clear to me who within the industry would be responsible for that,’ he said. The ideal would be a single multi-purpose GNSS-based system capable of dealing with everything from signalling to track maintenance.
While warning that GNSS ‘isn’t going to solve all your problems’ Cross said that it could play a major role if used in conjunction withother positioning technologies.
He said a ‘meeting of minds’ between the industry and representatives of other transport sectors such as the Civil Aviation Authority could be fruitful.
Prof Frank Kelly, chief scientific adviser to the Department for Transport, underlined the importance to the rail industry’s future of successfully deploying new technology.
He said GNSS could play a part in helping the railways fight back against other modes of transport, which have shown considerable success in adopting new technologies rapidly and effectively.
Kelly pointed out that in the European aviation sector, air traffic control delays have plunged and fares have halved. In the car industry, maintenance problems have fallen and vehicle prices have dropped.
‘Competitor transport industries seem to make quite rapid progress,’ he said.