Papers
4
Total Citations
409
H-Index
3
About
Deegan Atha is a robotics and computer vision researcher whose work spans structural health monitoring, autonomous off-road navigation, and planetary exploration. He is perhaps best known for his 2017 study on deep learning-based corrosion detection, which has accumulated an impressive 362 citations and demonstrated how convolutional neural networks could replace time-consuming, subjective visual inspections of infrastructure — a contribution with significant real-world implications for safety and economic efficiency. This work helped establish neural networks as a viable tool in automated structural assessment. More recently, Atha has turned his attention to the challenges of high-speed autonomous navigation in unstructured outdoor environments. His RoadRunner series of papers addresses traversability estimation — teaching robots to intelligently interpret complex terrain geometry and perceptual conditions using onboard sensing alone. The follow-up RoadRunner M&M extended this framework to multi-range, multi-resolution mapping, pushing the boundaries of what autonomous systems can achieve at speed. Alongside this, his research into semantic mapping for planetary robotic explorers tackles the critical problem of precise localization in GPS-denied extraterrestrial environments. Together, Atha's body of work reflects a researcher committed to making autonomous systems safer, smarter, and deployable in some of the world's most demanding conditions.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4