SPEAKING at IRJ’s International Railway Summit held with IRITS Events in Paris, Professor Andy Doherty, Network Rail’s (NR) chief technology officer, started his presentation by explaining that both the Shift2Rail research programme and NR have broadly the same objectives: to double capacity and energy efficiency and improve performance by halving failures and costs. Doherty said this would “radically change the railway travel proposition to the passenger and freight forwarder” and improve competitiveness.


Doherty outlined the situation in Britain and the challenges it faces. “Passenger numbers have doubled since 1997-98, the first full year of rail franchising,” he said. “Individual incidents of delay fell by 40% between 2006-07 and 2014-15. However, the impact of each delay has grown as the railway has become more congested. We need more solutions to increase capacity.

AndyD“NR’s operating costs have fallen by 46% since 2003-04,” Doherty continued. “This reduction is primarily due to the introduction of new technology, innovation and in-sourcing some key activities.”

Doherty sees a future driven by innovation where technology focuses on areas where there is a strong need for research and development rather than deployment.

“We need a digital intelligent train,” Doherty said. “It would have composite/metal bodyshells created through 3D printing and adaptive manufacture, with common European traction packages and running gear configurable for high-speed or suburban operation. Initially the digital train would have mechatronic bogies, but ultimately these would be replaced by independent self-steering low-damage wheel units. An integrated braking system with all-electric service brakes would be necessary for high-capacity train convoying.

Continuous communications, mobile and data services would enable each train to become a ‘gateway’ for all passenger, train control, and train maintenance systems. The train would have a self-learning and adaptive train control system to match ETCS/ATO needs, with ultimately convoy and virtual coupling capability. Self-diagnostics with fully-integrated 4.0 monitoring and status systems for both the train and the infrastructure would be necessary for predictive maintenance.

“But perhaps the first Shift2Rail train could be a tram-train?” Doherty mused. “For lines with signalling, a true low-cost ETCS needs to be developed, that is purely ATP, using distance-to-go with on-board sensors and satellite navigation, and line-of-sight control on non-signalled lines.”

Doherty says the bodyshells and structures would be based on bus technology with front and rear crumple zones and common seats and fixings, but not like Britain’s infamous Pacer DMUs which mounted a bus body on a freight wagon underframe.

Single motored wheel units with independent suspension would be used instead of bogies. Wheel motor units are already being developed with an output of 5kW, and at 20kW in the near future, while mechatronics removes the need for axles. Electric and intelligent magnetic/eddy-current brakes will remove the need for any air systems.
Doherty said railways need to introduce convoying to make better use of capacity. He said the first step is to develop all the functions and capabilities of ETCS Level 2 with driver advisory systems, and automation to GOA2 initially and then GO4.

“It is essential to achieve a viable business case,” Doherty said, “so we need a radically simplified interlocking philosophy where design and functional activities are separated with minimised SIL 4 operation.

“We need to junk the overlap rules, tear up the rule book and rethink how we operate the railway.”

He says ETCS Level 3 functionality and possibly beyond to Level 4 would require train-to-train communications, signal overlap rules based on assured braking performance, control systems adapted for different train capabilities, with train-assured braking capability.

Train convoying will become possible with self-steering and virtual coupling of trains, and robust secondary safety features, such as derailment containment and on-board obstacle detection.
Doherty said convoying, better traffic management, automatic digital trains, and advanced communications and ticketing will provide the basis for new service concepts.

Track redesign

Doherty says the track and sub-layer account for 50% of NR’s costs while points make up 20% of failures. He believes adaptive mechatronic points could reduce switches and crossings dramatically. “We have the opportunity to rethink rail design and metallurgy, with self-healing and lubricating rail steel nanotechnologies, 3D and ‘additive’ manufacturing of rail crossing and rail parts,” Doherty said.

“The technologies and vision described above come with a price,” Doherty observed. “Railways will need to develop the skills and competence to specify, operate and maintain the technologies. They need to rapidly achieve expertise in specifying, managing, and operating complex software systems.”

Doherty said standards will need to be developed for many of these technologies while railway system and product authorisation and verification must move to an aerospace design verification process.

“If we do everything in Shift2Rail, do we really need Hyperloop?” Doherty concluded.