Max Robotics — Global Service Robot Directory + US Certification

About

Jeremy Paul Coffelt is a researcher advancing the frontier of autonomous underwater robotics, with key contributions in computer vision, perception, and mission explainability. His work addresses fundamental challenges in marine environments—particularly the degradation of sensor data caused by turbidity, light attenuation, and marine snow. Coffelt’s most cited paper, “A Deep Learning Framework for Semantic Segmentation of Underwater Environments” (2022, 7 citations), introduces a neural network architecture that enables robots to classify and segment underwater scenes for tasks like bathymetric surveys and infrastructure inspections. In “Marine Snow Simulation and Elimination in Video” (2023, 5 citations), he tackles the pervasive problem of organic debris that compromises visual algorithms, proposing both a simulation pipeline and a removal method. His recent work, “Implementation and Application of a Knowledge Service for AUV Mission Explainability” (2025, 4 citations), pioneers the use of large language models and graph databases to allow natural-language querying of autonomous vehicle decisions—a step toward transparent, trustworthy robotics. Coffelt also developed SAVOR (2024, 3 citations), a sonar-aided visual odometry system that fuses imaging and acoustic data to overcome scale ambiguity and poor visibility. With a growing citation record and a focus on practical, deployable solutions, Coffelt is shaping how underwater robots see, understand, and explain their world.