Project number
27052
Organization
Universal Avionics
Offering
ENGR498-F2026-S2027
Helicopters operate low and slow in cluttered environments where wires, towers, poles, cranes, and terrain are the leading causes of controlled-flight-into-obstacle accidents. A large amount of obstacle information already exists in open-source geospatial databases and open imagery sources, but it is fragmented, inconsistently formatted, and not directly consumable by airborne avionics. This project asks whether those open sources can be aggregated and normalized into a single obstacle database in a standard, avionics-consumable format (target: ARINC 815), and whether a vision system can both validate the data against real imagery and visually augment (overlay) detected obstacles onto that imagery to improve pilot obstacle awareness and detection.
Scope
1. Work with UA engineers (avionics data, vision/imaging, and software) to understand the ARINC 815 format and how airborne systems consume it, plus the accuracy, integrity, and update requirements that matter for flight use.
2. Survey and evaluate available open-source obstacle and geospatial databases and open imagery sources for coverage, accuracy, currency, and licensing; determine what is usable and where gaps exist.
3. Design a pipeline that ingests these open sources, deconflicts and normalizes them, and outputs a consumable obstacle database in ARINC 815 format.
4. Validate the dataset using Google Maps / open aerial and street-level imagery to confirm obstacle presence and attributes.
5. Develop a prototype vision system to detect obstacles in imagery, correlate detections against the database, and artificially paint/overlay the known obstacles onto the image to demonstrate improved detection and situational awareness.
6. Test with sample geographies supplied or agreed with UA, measure detection performance and data quality, and iterate.
7. Develop a plan (including cost, data-licensing, and scalability considerations, and a path toward the data integrity/certification questions UA would need to address) to move from prototype to a usable tool.
8. Present results in a video conference and PowerPoint presentation to a group of UA engineers.
Additional Resources
UA will supply guidance on the ARINC 815 format and consumption requirements and can provide sample areas or reference data as agreed.
Skills Requested
Software engineering, data engineering/ETL, GIS and geospatial data, computer vision and image analysis, machine learning, data analysis and visualization. Bonus knowledge: avionics/aviation data standards, remote sensing/photogrammetry, systems engineering, statistics.
Benefits to Society
Better obstacle detection for rotorcraft saves lives by protecting crews and the public during emergency medical services (EMS/air ambulance), search and rescue missions, law enforcement activities, firefighting, and utility/infrastructure operations, where low-altitude flight near wires and towers carries the highest risk.
Scope
1. Work with UA engineers (avionics data, vision/imaging, and software) to understand the ARINC 815 format and how airborne systems consume it, plus the accuracy, integrity, and update requirements that matter for flight use.
2. Survey and evaluate available open-source obstacle and geospatial databases and open imagery sources for coverage, accuracy, currency, and licensing; determine what is usable and where gaps exist.
3. Design a pipeline that ingests these open sources, deconflicts and normalizes them, and outputs a consumable obstacle database in ARINC 815 format.
4. Validate the dataset using Google Maps / open aerial and street-level imagery to confirm obstacle presence and attributes.
5. Develop a prototype vision system to detect obstacles in imagery, correlate detections against the database, and artificially paint/overlay the known obstacles onto the image to demonstrate improved detection and situational awareness.
6. Test with sample geographies supplied or agreed with UA, measure detection performance and data quality, and iterate.
7. Develop a plan (including cost, data-licensing, and scalability considerations, and a path toward the data integrity/certification questions UA would need to address) to move from prototype to a usable tool.
8. Present results in a video conference and PowerPoint presentation to a group of UA engineers.
Additional Resources
UA will supply guidance on the ARINC 815 format and consumption requirements and can provide sample areas or reference data as agreed.
Skills Requested
Software engineering, data engineering/ETL, GIS and geospatial data, computer vision and image analysis, machine learning, data analysis and visualization. Bonus knowledge: avionics/aviation data standards, remote sensing/photogrammetry, systems engineering, statistics.
Benefits to Society
Better obstacle detection for rotorcraft saves lives by protecting crews and the public during emergency medical services (EMS/air ambulance), search and rescue missions, law enforcement activities, firefighting, and utility/infrastructure operations, where low-altitude flight near wires and towers carries the highest risk.