by Martin Litzenberger, Carmina Coronel and Kilian Wohlleben (AIT Austrian Institute of Technology GmbH)
Accurate real-time traffic sensing is of key importance for optimising and managing road traffic. Often the high density of traffic sensors, needed to achieve an accurate real-time monitoring of important roads, is difficult to implement due to technical constraints or because of high installation cost. Fibre-optic sensing (FOS) is a new and cost-effective alternative technology that allows a seamless, real-time monitoring of the road traffic over large distances of up to 50 km, even in remote areas such as on critical costal or mountain roads, using existing telecom fibre-optic cable infrastructure.
Even with an upcoming transition to electric mobility and a modal shift from individual to public transport in the future, roads will stay the backbone of transportation for years to come. Therefore, permanent traffic monitoring is crucial to ensure optimal traffic flow. The data provided by real-time road-traffic monitoring provides information regarding traffic jams or accidents. With such information, traffic-management centres are enabled and supported to react quickly to incidents and intelligent transportation system (ITS) measures, for example, the closure of a lane or temporary usage of the hard shoulder, can be imposed. Often the large number of traffic sensors, needed to achieve an accurate real-time monitoring of important roads, is difficult to implement due to technical constraints or because of high installation cost. Furthermore, in remote areas where traffic monitoring can be also of importance (e.g. on difficult and narrow road sections in costal or mountain roads), a dense coverage and easy connectivity of road-traffic sensors is problematic. One method for traffic monitoring is through crowdsourcing of smartphone connection data [1] or from fleets of vehicles equipped with GPS systems (“floating car”). Google Maps is the most prominant example of the crowdsourcing approach. However, it is unable to deliver true real-time information, it relies on traffic models, and it needs the “cooperation” of the data providers, that is, a high enough number of mobile phone users in the area of interest. In addtion to that, mobile phone and wireless data connectivity might also be reduced in remote areas.
Fibre-optic sensing (FOS), also often termed “distributed acoustic sensing” (DAS) [2], is a technology that allows a seamless, real-time monitoring of vehicle trajectories on a road over large distances of up to 50 km without additional roadside installations. It uses one unused optical fibre of an optical cable already installed in the ground for data and communication networks (telephone, Internet), as a distributed sensor. With increasing “fibre-to-the-home” initiatives througout Europe, even in remote areas, fibre-cable infrastructure will become more availble, and will be typically installed next to roads.
In FOS systems, an interrogator device connected to one end of the fibre transmits a series of laser light pulses into the fibre-optic cable, then portions of the light pulses are back-scattered from inside the fibre and are measured by the interrogator. Any vibrations generated by road traffic nearby the cable stretch and compress the optical fibre on a microscopic scale, affecting the optical-path length. An interferometer in the interrogator unit measures time of flight and the phase shift of the back-scattered light and thereby determines the position and trajectories of road vehicles (Figure 1a).
We have developend and applied a FOS traffic-flow-sensing algorithm [3] over a distance of 19 km with the fibre-optic cable being installed in parallel to a highway in a mountain region in Austria, and compared the results to the speed measurements from a reference induction-loop detector installed in the road surface of the same road.
The FOS signal is a combination of signals originating from passenger vehicles and trucks for both lanes of the highway. The current detection algorithm does not distinguish between vehicle classes nor lanes. In line with this, we performed the comparison by comparing our results to both accumulated truck and passenger vehicle results from the reference detector. The box plots (Figure 1b) show the average speeds of the passenger cars (ground truth) and trucks (ground truth) in 60 one-minute intervals, compared to the average speeds estimated by the FOS algorithm, in the same time intervals, at the same position along the road.
Figure 1: (a) Principle of FOS road-traffic monitoring. The fibre optic cable picks up vehicle vibrations that are probed via the back-scattered light of the interrogating laser pulse. (b) Comparison of induction-loop counter-reference-speed measurements (cars and trucks separated) and the accumulated average speed estimated from the FOS measurement (no vehicle classification).
The reference detector shows that the speed of the measured trucks (n = 94) is in the range of 80–100 km/h, which is slower than the passenger vehicles (100–135 km/h). The estimated average speeds extracted by the FOS algorithm show that the FOS results are in good agreement with the average speed of trucks, suggesting that the cars are not picked up by the FOS measurement.
Passenger cars and vehicles on the opposite carriageway, away from the fibre-optic cable, were not detected reliably. In detail, we found that: (i) passenger cars were not detected if the signal intensities were not strong enough, because the roads’ traffic lanes were too far away from the fibre cable (up to 20 m distance in the described setup), (ii) the algorithm failed when vehicles were driving exactly in parallel or very close to each other, thus preventing our algorithm from correctly identifying the two distinct trajectories, or (ii) weaker signals originating from passenger vehicles have been masked by stronger signals originating from heavier and larger vehicles.
In conclusion, we found that the FOS system is capable of monitoring traffic situations based on average speed measurement but cannot reliably resolve smaller passenger cars. Also, the distance of the fibre-optic cable from the traffic lanes is crucial, and thus not all lanes of wide roads can be monitored. Therefore, we see the potential application of FOS traffic monitoring on, for example, smaller roads in remote areas, such as on costal or mountain roads, where a dense coverage and easy connectivity of conventional road-traffic sensors is problematic.
References:
[1] M. Lewandowski, et al., “Road traffic monitoring system based on mobile devices and bluetooth low energy beacons,” Wireless Communications and Mobile Computing, vol. 2018, pp. 1–12, Jul. 2018. [Online]. Available: https://doi.org/10.1155/2018/3251598
[2] D. Hill, “Distributed acoustic sensing (DAS): theory and applications,” in Frontiers in Optics, 2015.
[3] C. Wiesmeyr, et al. “Distributed acoustic sensing for vehicle speed and traffic flow estimation”, in 2021 IEEE international intelligent transportation systems conference (ITSC), 2596–2601, 2021. Available at: https://doi.org/10.1109/ITSC48978.2021.9564517
Please contact:
Martin Litzenberger, AIT Austrian Institute of Technology GmbH, Austria
