The microscopic graph twin of the UK rail network.
A unified topological model that fuses physical infrastructure, low-level safety interlocking logic, and real-time signal-box telemetry — engineered to dissolve the £100m+ sub-threshold delay sink invisible to TRUST.
Up to 40% of delay minutes are invisible to the systems that run the railway.
UK rail measures performance through destination-based metrics (PPM) and manages delays through TRUST. But Swedish and wider European research confirms that delays below the 3-minute reporting threshold cascade to drive massive secondary delay across the network — and are completely unallocated.
The MGT calculates performance variance edge-by-edge across the graph:
SRTᵢ = planned Sectional Running Time (BPLAN)
Flag when 0 < δᵢ < θTRUST
Three intractable problems. One graph.
Isolate the minor, recurring delays under 3 minutes that fall below TRUST reporting but propagate into massive secondary network delay. £15m+ of unallocated delay minutes on the Wessex route alone.
Reverse-decode real-time signal box logs — from Solid State Interlocking down to legacy mechanical boxes — to verify that every actual train movement matches the strict criteria of the signalling control tables.
Feed micro-operational performance data directly back into the Integrated Train Planning System (ITPS) to dynamically optimise sectional running times and resolve chronic bottlenecks before they reach the timetable.
Direct delivery against GBRX core priorities.
Converting physical layouts and control tables into machine-readable logical models accelerates automated safety cases and digital interlocking verification.
Eliminating chronic sub-threshold bottlenecks is a low-cost, software-driven capacity increase — avoiding billions in physical infrastructure spend.
Digesting manual logs and mechanical token exchanges ensures legacy, rural, and heritage routes are not left as digital blind spots in the transition to GBR.