Real-world AI & field operations for nature protection.
Date & place: Thu 5th March 2026 as part of the Nature Drone Forum – Nairobi, Kenya.
Teams work on standardised conservation datasets and (for selected tracks) short, safety-managed flight windows. Findings are shared openly so others can learn and reuse.
Why DUCC?
A one-day, multi-track comparative challenge that tests impact, technical performance, and operational readiness of current tools and teams—so practitioners can adopt what works. Outputs include public track briefs and an overview report (CC BY 4.0), while teams keep their own IP.
AI-first tracks (data provided)
- Thermal wildlife search & localisation (precision/recall on unseen sequences)
- Wire-snare detection in woodland imagery
- Habitat mapping / semantic segmentation from multi-/RGB imagery
- Through-water sensing (water clarity permitting): object/feature detection
Other tracks
- “Drone-mule” logistics planner (route/energy constraints; sim data)
- Safe photography / disturbance-aware capture (policy + model prompts
Each track supplies: training samples, a held-out test set, and a clear metric (e.g., mAP/IoU/F1 + runtime). Some tracks are data-only (remote); others include short, supervised flight windows for on-site scoring.
What teams receive
- A brief & rules for the chosen track, plus baseline scripts/format
- Standardised datasets (train/validation) and a blind test set for scoring
- A submission template (model/code + 2-page methods note)
Judging & fairness
- Scoring rubric: Impact (30) / Technical (30) / Operational Readiness (30) / Inclusion & Ethics (10).
- Conflicts declared and managed; judge panels mix practitioners, ops leads, and community reps.
- No AI is used to judge—AI may help with grammar/layout only.
Safety & ethics
Trained marshals manage RAMS and wildlife buffers; code of conduct applies. Public outputs avoid sensitive locations; media by consent only.
Outcomes
- Recognition awards for innovation & ethics and a live finalist showcase during NDF.
- Adoptable prototypes & guidance that feed straight into conservation practice.
- Track briefs online by 31 March 2026.
Who should enter?
University teams, startups, NGOs, tech providers, and field units working at the intersection of AI, drones, and conservation operations. Partnerships across research and practice are encouraged.
Timeline (high level)
- Now: Team registration opens (limited slots per track)
- Pre-event: Datasets & rules released to accepted teams
- 5 March 2026 during Nature Drone Forum: Global Drones in Nature Conservation Symposium: On-site evaluations & finalist showcase
- By 31 March 2026: Public briefs/reports published.
Ready to join?