Abstract—Most urban intersections still rely on fixed-cycle traffic signals—a setup that works well enough under predictable conditions but falls short when queues shift unexpectedly or an ambulance needs to get through fast. This paper presents SmartSignal AI, a full-stack TypeScript platform that combines pressure-based adaptive signal control, software-level emergency preemption, and short-horizon congestion forecasting. The system uses a 250ms control tick and models a single fourapproach intersection with Poisson vehicle arrivals and a heterogeneous fleet consisting of cars (62%), motorcycles (18%), trucks (10%), buses (8%), and priority vehicles (2%). All testing was done in simulation: a lightweight SUMO-inspired microsimulation kernel written in TypeScript that replays the same deterministic demand trace to both the adaptive and fixed-cycle controllers under identical conditions. Across five nominal perapproach demand levels from 300 to 1,500veh/h, the adaptive controller reduced average waiting time and queue length by 64.3% at 300veh/h; at 600–1,500veh/h, the reductions were 1.7–20.1% as both controllers approached saturation. Throughput gains remained between 3.1% and 4.7%. Emergency events wereinjected synthetically, and each event was processed within one 250ms simulation tick. The forecasting subsystem uses a 120s horizon. We do not report numerical forecasting accuracy in this paper—the codebase does not yet include the logging infrastructure needed to pair each prediction with what actually occurred, so any accuracy figure would not be reproducible. Everything reported here comes from simulation; we have not tested the system on physical hardware or at a real intersection. Index Terms—Adaptive traffic signal control, Max-Pressure, intelligent transportation systems, emergency vehicle preemption, congestion forecasting, micro-simulation, real-time control.

Adaptive Traffic Signal Control; Max-Pressure Control; Intelligent Transportation Systems; Emergency Vehicle Preemption; Congestion Forecasting; Traffic Microsimulation; Real-Time Traffic Control.