Stoke Junction Case Study

The project began in May 2023 as a six-month trial spearheaded by London North Eastern Railway (LNER) and Network Rail, initially utilizing over 250 first-generation DeterCam units to address critical deer incursions.

Following the immediate success of the trial, the programme transitioned to a commercial rollout under Network Rail’s management, expanding across the Stoke Junction operational sector.

High deer populations and established crossing paths created severe collision risks, rolling stock damage, and service disruptions, with an historical average of 15 deer strikes per year at Stoke Junction alone. Site inspections revealed frequent deer tracks, droppings, and significant gaps in trackside vegetation and rotted boundary fencing.

Implementing traditional exclusionary fencing presented major financial and logistical hurdles, particularly where boundaries bordered third-party agricultural land. Furthermore, extensive fencing restricted natural wildlife movement and remained vulnerable to ongoing damage, necessitating a reliable, non-lethal automated alternative.

DeterCam units were strategically positioned at identified crossing points, wooded gaps, and open boundaries. The deployment evolved across sequential phases to adapt to environmental demands and operational findings:

• Phase 1 (Stoke Junction Trial):
A 2-mile deployment using 200 devices equipped with solar panels, rechargeable batteries, and Wi-Fi routers. AI algorithms were trained to ignore non-target species (like foxes, badgers, and domestic pets) and strictly target deer.

• Phase 2 (System Enhancement):
Recognizing that UK winter sunlight limited solar efficiency, sensors were upgraded with larger internal batteries and 4G communication modules. Three site visits were conducted to recalibrate camera angles and reinforce damaged fence lines.

• Phase 3:
Installed v2 at Stoke Junction.

• Phase 4:
Installed v3 at Stoke Junction.

• Phase 5:
Install v4 at Stoke Junction

The core objective is to prevent deer–train collisions, minimize associated service delays, and offer a cost-effective, non-lethal infrastructure protection model. Additionally, the system functions as a robust biodiversity monitoring tool, logging species data, habitat usage, and behavioral responses for Network Rail’s environmental tracking.

• 100% Collision Reduction: Zero deer strikes have been reported within the protected areas during their monitored deployments since May 2023 (down from the baseline of 15 per year).
• High-Volume Deterrence: Over 13,200 This figure till (11th October 2026) deer deterrence events have been recorded across the program to date.
• Immediate Impact: The initial Stoke Junction activation deterred 56 deer on its very first day, clearing animals within four hours of going live.
• Phase Success Rates: The system deterred 680 deer during Phase 1, and over 2,500 deer during Phase 2. At peak crossing zones, deterrence activity frequently exceeded 100 deer in a single day.
• Behavioral Insights: The system tested over 20,000 unique deterrence combinations (visual lights, audible sounds, and ultrasonic frequencies), automatically labeling which alarm profiles were most effective for specific deer species.

Continuous monitoring will drive further enhancements to AI detection models, deterrence efficacy, and hardware reliability. The extensive data collected will help Network Rail identify emerging wildlife crossing hotspots, proactively guide future installations, and support long-term biodiversity and railway safety objectives.