POC Report: Isis at Brazilian Air Force EXOP Carranca 2025
Proof of Concept for Isis Software in Real Exercises
Introduction to Isis
Isis was designed by professionals with extensive experience in SAR missions, both on land and at sea. The multidisciplinary team brings together specialists active in all stages of the process, from planning to the execution of the rescue action.
The project was born from the goal of incorporating the most current theories of search effort optimization, combining mathematical evolution with recent computational capabilities to enable more efficient, precise planning consistent with modern Search and Rescue System demands.
The methods that prevailed until then, developed to allow manual calculations, made predictions based on average values, which reduced accuracy. Isis brings modern solutions to these limitations.
Technical Innovations
Monte Carlo Simulation
Instead of projecting single drifts, Isis creates areas representing uncertainties through thousands of particles simulated over the sea in Monte Carlo progressions, each one 'living' the same physical environment the target would experience.
Probabilistic Heat Maps
These particles, moving under gradual diffusion in one-hour steps, produce physically plausible trajectories and generate a more accurate probabilistic map (POC), resulting in highly representative heat maps.
Drift Models
- Point - When LKP is a known point
- Trajectory - For uncertainty line between two points
- Course & Distance - For heading and distance uncertainty
Bayesian Update
After each sortie, results are accounted for and the possibility scenario is readjusted. The algorithm automatically maintains the computation of the effects of searches performed on subsequent sorties.
Test Scenarios - EXOP Carranca
Florianopolis, SC - Brazilian Air Force
Scenario 1 - PIW (+10h)
Blue marker represents the LKP. After 10 hours of maritime drift, Isis automatically generated a probabilistic heat map.
Spatial alignment consistent with local winds and currents.
Scenario 2 - PIW (+23h)
With the PIW in drift for 23 hours, Isis recalculated the probability distribution based on updated history.
Efficient Task Area to maximize POS with only 2 flight hours.
Scenario 3 - 10-person Raft (+30h)
Survival raft for 10 people remained adrift for 30 hours.
Suggested Task Area for 3 flight hours, prioritizing highest POC density region.
Real Case - Land Search
Aircraft Not Located by FAB
Real search mission in land environment involving an aircraft with two people on board. Initial planning was conducted by FAB, but the aircraft was not located during aerial operations.
The methods employed at the time provided only one overflight over the wreckage. The IAMSAR format could have provided greater probability of location.
Aggravating Factors:
- Large initial area
- Slow replanning
- Geography with vegetation and shadows (high masking factor)
- Pattern edges with less observer attention
- Lack of probabilistic prioritization
Results Comparison
| Description | Real Mission | Planned with Isis | Efficiency |
|---|---|---|---|
| Hours to 1st overflight | 32:30 | 5:32 | -83% |
| Cost (SC-105) | US$ 226,395 | US$ 38,313 | -US$ 188k |
| Overflights over wreckage | 1 | 6 | +500% |
| POS | ~15% | ~92% | +513% |
Results and Conclusions
The impacts resulting from incomplete planning in search missions - where there are numerous uncertainties - tend to intensify due to the scalar growth of errors.
Modern mathematical methods, combined with consistent logical processing, support decisions much more efficiently.
This is the answer that Isis delivers.
Transform Your SAR Operations
Discover how Isis software can optimize your Search and Rescue missions.