The challenge

Background and context

CAF operations generate vast, heterogeneous data streams such as text reports, imagery, video, audio, Radio Frequency (RF)/signal intelligence, sensor telemetry, and emerging modalities such as quantum-derived measurements and drone-based information. These data sources remain siloed, limiting real-time situational awareness and decision-making. 

This capability directly supports the CAF Digital Campaign Plan and the DND/CAF AI Strategy by enabling learned, adaptive fusion across modalities rather than static aggregation. It aligns with Force Capability Plan priorities for Intelligence Surveillance Recognition (ISR), Command & Control (C2), and operational resilience. The capability will enhance CAF’s ability to integrate intelligence across domains and classification levels, ensuring interoperability with allies and secure operations in contested environments. 

Unlike traditional rule-based fusion, this initiative leverages AI-driven architectures to learn complex relationships across heterogeneous modalities, propagate uncertainty, and deliver policy-aware, explainable outputs for mission-critical decisions in dynamic, degraded environments. 

Solutions will be integrated into CAF ISR platforms, C2 systems, and tactical edge deployments. Outputs will inform CAF doctrine for multi-domain operations, feed into experimentation and wargaming environments, and support procurement strategies for next-generation decision-support systems. 

DND/CAF are hosting this challenge to observe advancements in AI technology to resolve this issue.

Examples of application can be described as, but are not limited to: 

• Joint ISR Fusion for Arctic Operations: AI-driven spatiotemporal alignment of satellite imagery, RF signals, and telemetry for persistent Arctic domain awareness. 
• Real-Time Threat Assessment in Multi-Domain Battlespace: Deep learning-based fusion of Electro-Optics (EO) video, Signals Intelligence (SIGINT), and text intelligence for dynamic targeting and force protection. 
• Edge Fusion for Tactical Units: Deploy constraint-aware AI models on wearable systems to integrate audio, video, and sensor data for soldier situational awareness under degraded connectivity. 
• Maritime Task Group Operations: AI-powered anomaly detection using sonar, RF, and visual feeds with uncertainty scoring and explainable outputs. 
• Airborne Multi-Sensor Platforms: Fuse radar, Electro-Optics/Infra-red (EO/IR), and telemetry for enhanced detection and tracking of stealth or spoofed adversary assets. 

Essential outcomes

Proposed solutions must demonstrate the following:

• Deliver an AI model that can aggregate, ingest, fuse, and generate outputs from at least two (2) heterogenous data types (e.g. sensor, text, RF) to produce output metrics and measures (e.g. classifications, detection, correlations)

Desired outcomes

Proposed solutions should include capabilities and considerations such as, but not limited to, the following:

• Advanced deep learning architectures for spatiotemporal alignment, uncertainty propagation, and confidence scoring across modalities 
• Entity resolution and dynamic knowledge graph integration for persistent object tracking across domains
• Policy-aware fusion leveraging AI-based provenance tracking for secure integration across classification levels with full lineage 
• Scalable architecture for real-time AI-powered fusion pipelines in operational environments, including explainable outputs for operator trust 
• Incorporate Size/Weight/Power (SWaP) and compute limits into fusion pipelines for edge deployment