Idea
A real-time thermal imaging and deep learning platform that detects deer and other animals and warns drivers to prevent collisions in connected vehicles.
Research Paper
Core Innovation
This paper introduces a system that integrates thermal sensing with deep learning models trained on a large custom dataset to detect deer in real time. It uniquely combines this with vehicle-to-everything communication to broadcast warnings with under 100 milliseconds latency. This approach outperforms traditional visible light cameras, especially in poor weather, enabling more reliable detection and timely alerts.
Market Size (TAM)
$2–10B TAM for automotive safety and collision avoidance systems; $1–2B SAM from connected vehicle manufacturers and fleet operators. Driven by increasing demand for advanced driver assistance and vehicle-to-everything communication technologies.
Potential Customers & Pain Points
- Automotive manufacturers seeking advanced safety features
- Fleet operators reducing accident-related costs
- Insurance companies lowering claims from wildlife collisions
- Government agencies aiming to improve road safety
- Wildlife conservation groups mitigating deer population decline
Business Model
Licensing the detection and alert platform to automotive OEMs and fleet management companies; offering software updates and data services for continuous improvement.
Competitive Landscape
- Mobileye
- Waymo
- NVIDIA
Implementation Challenges
- Integration with diverse vehicle platforms
- Regulatory approvals for safety systems
- Adoption of CV2X communication infrastructure
Validation Strategy
- Conduct extended field trials across varied geographic regions
- Explore adding additional animals apart from deers in detection system
- Partner with automotive OEMs for pilot integration
- Collect and analyze real-world collision reduction data
Research Paper Overview
Real-time Deer Detection and Warning in Connected Vehicles via Thermal Sensing and Deep Learning
Summary
This paper presents a real-time system combining thermal imaging, deep learning, and vehicle-to-everything communication to detect deer and warn drivers, aiming to reduce deer-vehicle collisions. The system was trained on over 12,000 thermal images and achieved high accuracy and low latency in field tests, outperforming visible light cameras especially in challenging weather. It broadcasts alerts to nearby vehicles and roadside units when deer are detected with high confidence, enabling timely driver warnings and enhanced road safety.