Sensors and Connectivity Series: How can LiDAR provide us with a new view of the railways?
Latest Update: October 2019
The use of LiDAR has expanded in the rail industry. In Denmark, an engineering company is mapping the rail network using LiDAR to produce a digital twin. In the UK, LiDAR systems mounted on passenger trains are being rolled out to survey the rail corridor.
What is LiDAR?
LiDAR (Light Detection and Ranging) is a remote sensing technology that generates images by reflecting laser pulses off surfaces. LiDAR is somewhat similar to radar. It works by using the time delay between the pulse being transmitted and it being received to determine the distance between the source and the target. While radar uses radio waves and microwaves, LiDAR uses ultraviolet, visible and (more commonly) near-infrared light. The advantage is that LiDAR can generate images with a greater resolution due to the shorter electromagnetic wavelengths used. Such imagery can be used to produce 3D maps and models of surfaces, as well as detecting the location and velocity of objects.
What industries use LiDAR?
LiDAR technology was first used in meteorology, to study clouds and pollution. Its application in meteorology has expanded and can now include determining wind speed and temperature at altitudes of up to 120 km from the earth’s surface.
Maps generated by aerial imaging have a wide range of applications. One such example is that following the 2010 Haiti earthquake, the U.S. military used a single jet to fly over the city of Port-au-Prince at 3,000m. In one sweep, an area of 600m2 was captured at a resolution of 30cm. This meant that the precise height of rubble could be conveyed to disaster relief teams.
In agriculture, LiDAR maps offer more information than conventional mapping. Machine learning has been applied to LiDAR imaging to map plants based on species, allowing weeds to be identified and treated by a robot. Furthermore, the technology can monitor the behaviour of flying insects, making the use of pesticides more efficient.
Driverless cars use LiDAR to map objects and obstacles in the car’s surroundings. Modern LiDAR systems can differentiate between a cyclist and a pedestrian in less time than a human can, potentially reducing motor-related incidents caused by human drivers.
Within archaeology, LiDAR has been used to map the remnants of settlements and offered new insights that are difficult to identify through other means. After an ancient Mayan city was discovered in 2012, one of the lead professors claimed that ‘in 45 minutes of flying, the LiDAR team accomplished a decade’s worth of archaeological survey’.
In the entertainment industry LiDAR is used to aid CGI production. During the filming of HBO’s Game of Thrones, the city of Dubrovnik was mapped using LiDAR to provide a 3D model for the fictional capital King’s Landing, with visual effects superimposed on the existing city.
How has it impacted the rail industry?
LiDAR has been used widely to map rail networks. In 2014 Network Rail conducted an aerial survey of its 16,000 km network at an unprecedented resolution of 20cm. LiDAR was used to map the elevation of the land surrounding rail infrastructure, a process which would have previously been done by teams on the ground in high-risk, remote locations. Network Rail’s New Measurement Train is fitted with a LiDAR scanner to inspect the track for any anomalies.
Siemens demonstrated an autonomous tram in September 2018 which travelled along 6 kilometres of track in Potsdam, Germany, without the need for human intervention. The tram navigated using LiDAR in tandem with other sensors. LiDAR as a ranging technology is already common as part of a collision warning system on trams elsewhere in Germany.
In September 2019, it was announced that the engineering company Fugro has been awarded a contract to produce a digital twin of Denmark’s railways using LiDAR. The data will support track maintenance and assist construction and signalling teams. Fugro has also produced a unit which can be temporarily affixed to passenger trains to collect track and trackside data without interrupting service. Full use of this system is expected to save 1200 manhours and 200 hours of track possession.
What uncertainties remain?
LiDAR instrumentation can malfunction or fail in adverse weather conditions. The reason for this is that precipitation reflects or scatters the light emitted by sensors, causing issues for systems that navigate using LiDAR. Autonomous vehicles are the most prominent example of this, since they sometimes register falling rain or snow as obstacles to be avoided.
Even in good weather conditions, the collection of dirt and dust on the LiDAR lens can block light. Therefore, frequent maintenance is required to keep them clean.
What is the current state of R&D?
With a preference for solid-state designs, the next generation LiDAR systems in development are moving away from mechanical systems with moving parts, as these are subject to fatigue. However, this new technology does not scan the field but rather ‘flash’ to illuminate the surroundings in an instant, and so are less precise with a reduced range of object detection.
To overcome the intrinsic limitations to LiDAR navigation, multi-faceted sensor systems are being developed. By deploying LiDAR in conjunction with other technologies, such as machine vision, the sensing system will navigate more effectively. Algorithms are being developed to help artificial intelligence understand dynamic systems, such as a roadways, based on multiple inputs.
What should the rail industry do?
LiDAR is a valuable instrument for harvesting data and the rail industry should integrate this with systems such as Geographical Information System (GIS), to capitalise on the potentials for more targeted maintenance. This would offer benefits such as reducing the time spent performing trackside inspection and limiting disruption of services due to engineering works.
LiDAR, with its ability to sense distance and relative velocity, could be an important step in developing autonomous trains. Since this technology has already been introduced into the automotive industry, research may determine how much work would be required to adapt existing technology for a rail environment.