Optimal-Control Suggestion for Congestion on Freeways using Data Assimilation of Distributed Fiber-Optic Sensing

A team of Japanese researchers has developed a novel method to combat freeway congestion by combining distributed fiber-optic sensing (DFOS) with data assimilation (DA) and multi-objective optimization. The approach, detailed in a paper presented at the Transportation Research Board 105th Annual Meeting 2026, uses existing fiber-optic cables alongside roadways to monitor traffic flow continuously and without dead zones—a key limitation of conventional sensors that only provide intermittent data. This real-time monitoring feeds into a simulation system that estimates the effectiveness of various control scenarios, such as variable speed limits (VSL) and inflow control, as part of active transportation and demand management (ATDM).

The researchers validated their system on a Japanese freeway, testing two control strategies. When using VSL alone, throughput improved by 5-14%, but mean speed gains were modest (0-8%). However, combining VSL with inflow control yielded significant results: throughput increased by 10-15%, and mean speed improved by 20-30%. The paper also provides a method to uniquely determine the optimal control solution from a set of Pareto-optimal trade-offs, balancing throughput and mean speed. The authors emphasize the importance of proactive control before congestion occurs and balancing lane occupancy, suggesting that fiber-optic sensing could revolutionize real-time traffic management by enabling precise, state-dependent interventions.