All the shifts that we call “industrial revolutions” can be viewed as successive stages of increased automation. The first in the 18th century was enabled by machines, coal and steam; the second, in the first half of the 20th century by automated mass production lines, oil and electricity; and the third in the 1970s by electronics, computer science and information technology.
So, what next? Picture a world with ubiquitous, networked, big data-crunching, artificially intelligent machines and autonomous systems and you'll have an idea of the coming 4th industrial revolution, or "Industry 4.0". A profound transformation of the way in which we live, travel, work, produce and organise our economies.
Competitive pressure and the pace of automation
This vision may sound a bit futuristic, yet, the future is now. Fuelled by advances in nanotechnology, advanced materials, additive manufacturing, big data analytics, autonomous vehicles and the Internet of Things, to name but a few, automation is inexorably picking up pace in all industries, under the pressure for businesses to maintain their competitive advantage in globalised markets.
Transport and logistics are among the sectors primarily impacted: while Amazon has developed small package aerial delivery drones to curb rising shipment costs, US freight logistics companies are considering deploying driverless trucks in platoons, in a business where labour costs account for 75% of shipment costs and driverless trucks are estimated to be eight times more productive than human driven trucks and not affected by fatigue.
Drivers of automation in transport
Cost is not the only driver of automation, and autonomous systems need not be conceived from a purely substitutive perspective to human labour, but can bring many benefits in a complementary way. Firstly, they have become indispensable auxiliaries in activities that are too “dull, dirty or dangerous” for humans. We can think of tunnel exploration during construction work, inspection and maintenance on high bridges, highways and railway tracks, surveillance of cargoes as natural examples. Secondly, automation is also key to intelligent transport networks, because it allows for things such as operational optimisation, predictive maintenance, and enhanced personalised customer experience.
Air transport embracing automation
The technology for pilotless air travel is already in regular use with ‘fly-by-wire’ systems having eliminated many of the routine pilot functions of the past, such as setting courses and switching navigation radio frequencies. Auto-landing and auto-pilot have also been in place for decades, and control towers at airports are likely to become a thing of the past, with plans at London City airport to relay HD imagery to a remote site in Hampshire, for instance. NASA is currently studying the concept of single-pilot airline cockpits, with a first officer on the ground monitoring several flights at once.
Predictive maintenance has shown impressive results for aircrafts. Rolls-Royce tracks the status of thousands of engines operating worldwide, using onboard sensors and live satellite feeds and generating terabytes of data over the course of a flight, to predict when something might go wrong. This has significantly driven operational efficiency up and maintenance costs down.
Inspection drones, fitted with high definition video cameras, and lasers to scan the outside of the plane, are used by companies such as EasyJet to assist with routine fleet checks, cut down on inspection time, carry out remote checks when the engineers are not available on site, and to inspect aircrafts at height without having to set up a rig, which is safer.
Automated systems are also increasingly being deployed at airports to enhance the passenger experience. Tokyo’s Haneda Airport recently added some ‘robotic personnel’ at the security checkpoint and its train terminal. “Nao” the robot is linked to a digital feed. Travellers can ask it about arrivals and departures; Nao can interact and provide flight data, as well as weather forecasts and other information. At Geneva Airport in Switzerland, another robotic system, “Leo”, is able to scan boarding passes and check luggage.
Road transport is going driverless and busy working on platooning
Tesla, BMW, Ford, Volvo, and General Motors are competing fiercely for the future driverless cars market. In 2012, IEEE projected that up to 75 % of vehicles in the U.S. would be fully automated by 2040. A year later, Morgan Stanley was anticipating a much faster market penetration of autonomous cars by the mid-2020s. In the UK, autonomous car production has been forecasted by KPMG to go above 2 million units per year by 2030, with all vehicles produced being above “Autonomy Level 3”, i.e. with full transfer of safety critical functions (Autonomy Level 1 and 2 are limited to driver assist functions).
In 2017, Nissan began testing a driverless car on the roads of London. The car was guided by five radars, four lasers and 12 cameras, on top of a driver behind the wheels who could intervene in case of emergency.
New Tesla cars with auto-pilot mode, advanced collision detection systems and lower maintenance costs are already considered safer than traditional cars by insurance companies, with a smaller insurance premium. By 2040, it is estimated that all road accidents will drop by 80%. Meanwhile, issues around software reliability, data privacy, and cyber security are being actively taken care of.
Autonomous trucks and buses have also been trialled recently: Uber’s autonomous truck “Otto” made its first long distance test in October 2016, delivering 50,000 cans of beer via the Colorado highways from Fort Collins to Colorado Springs. The same year in Arizona, Local Motors embedded IBM’s “Watson Internet of Things for Automotive” technology into a driverless shuttle bus manufactured by 3D printing. “Olli” transported its first passengers on the streets of a shopping district just outside Washington D.C., and Local Motors is now planning to produce commercial cargo vehicles based on the same design.
Automated subway systems are on the rise
Metro lines have proven to be ideal environments to test and deploy driverless technology. The first completely automated subway train went into service in New York in 1961. Today, there are 55 fully automated metro lines in 37 cities around the world, operating 803km in total. The International Association of Public Transport (UITP) world report on metro automation estimates that by 2025 there will be 2,300km of automated metro lines in operation. It explains how this will allow operators to optimise the running time of trains, increase the average speed of the system, shorten headways up to 75 seconds, and reduce dwell time in stations to 15 seconds.
And what about automation on mainland railways?
Overground train networks present a higher challenge for driverless train technology, and necessitate more sensors, fences, obstacle detection capabilities, and vehicle-to-infrastructure communications (eg with level crossings), but some countries have already taken the lead in bridging this gap.
In Germany, Deutsche Bahn (DB) is working on bringing autonomous technology to its vehicles. A 30km section of track in Saxony near the Czech border has been constructed for test purposes for the Erzgebirgsbahn, and trials are underway foriron freight trains. The trains being tested by DB RegioNetz Verkehrs GmbH are making use of cameras and other collision-detection technologies.
In partnership with Prorail (Netherlands rail infrastructure firm), DB Cargo has started to trialdriverless freight trains on the 90-mile-long trackof the Betuweroute, between Rotterdam (Netherlands) and Emmerich (Germany).
Deutsche Bahn chairman Rüdiger Grube estimates in an interview with the Frankfurter Allgemeine Zeitung, that DB will have some fully automated trains in operation on their network by 2021-2023. In that interview, he explains that the rapid development of autonomous vehicle technology makes investment in driverless trains a necessity in order to remain competitive into the future.
In Western Australia, the global mining company Rio Tinto has developed a project called AutoHaul, currently in test mode, which aims to run an automated long-distance heavy-haul rail network, starting in 2018. Fitted with radar, sensory equipment and mapping technology, the autonomous machines can tell when an object is blocking their path and can respond to reduce the likelihood of impact.
The applications of automation in the railways go way beyond driverless operations.
In the US, GE has designed a Movement Planner which processes large amounts of data on the status of trains and track, to come up with near-optimal plans for moving trains in real time. GE advertises the possibility of realising a 10% velocity increase, i.e. 3 - 4 mph of increased speed per train. In North America, GE estimates that increasing the average speed by 1 mph on freight trains (av. 20 mph) saves a class 1 railroad $200 million per year. Big data analytics can lead to capacity improvement without laying additional track; for example, a railroad running 20 trains a day between NYC and Washington D.C. could increase that to 23 trains a day. Finally, the Movement Planner is the equivalent of a cruise control system for train drivers. The technology can assess the terrain and location of the train to calculate the optimal speed to run the locomotive for fuel economy.
The use of drones is increasing, and they have a greater role to play. Deutsche Bahn began using
miniature helicopter drones in Germany to combat graffiti-spraying gangs. In Poland, drones are routinely used to deter freight cargo thefts and pursue trespassers. Dutch railway company ProRail uses drones equipped with infrared sensors to check the switch point heating systems on its tracks. Union Pacific in the US is also looking at using drones for primary inspection and then sending humans to do another final inspection if something is found by the drones.
The more robots will be able to perform inspection and maintenance tasks on the tracks, the fewer risks human staff will be taking. A robot named Felix produced by the Loccioni Group in Italy is able to perform automatic inspection of rail switches. In Austria, a rail welding robot manufactured by Plasser & Theurer (APT 1500 R) has performed automated welding sequences without manual interaction, achieving a high level of welding quality.
Real-time, integrated and automated data flows also have the potential to radically enhance passenger experience. Providing real-time information to passengers and making their journey as seamless as possible might be achieved using an array of technologies, from digital tools such as journey planner mobile applications to intelligent ticket gates that use Wi-Fi beacon or facial recognition. The common principle is to provide the customer and the operator with the right information, at the right place and at the right time in an automated way.
We may see, in a not-too-far future, smart seats capable of knowing and remotely advertising their availability; apps giving passengers train crowding information; effortless and intelligent guidance through stations and on the platform; gateless access that avoids congestion, and smart ticketing systems that allow passengers to pay for their travel without pre-booking, and without even having to take a card out of their pocket. And intermodal services (mobility as a service) that allow passengers greater flexibility and connectivity, solving the “last mile” problem.
Delivering the intelligent railway networks of the future
In 1901, when Henry George Wells published hisAnticipations of the Reaction of Mechanical and Scientific Progress Upon Human Life and Thought, the UK rail industry was still at the edge of technological change. Today, the rail industry must embrace what the MIT technology review has called the “relentless pace of automation” if it wants to remain competitive, and take on the challenge of providing increasing value to its customers. Expectations are high, as passengers will want more and more detailed and up-to-date information, intermodal, reliable and punctual services, and all the added benefits of a digitised world. What are your thoughts? Will rail thrive or merely survive?
In the next article Vaibhav Puri, Head of Technical and Regulatory Policy, will look at the legislative and regulatory challenges associated with the introduction of truly autonomous systems on the railway.