Amanda Adkins

Papers

Total Citations

H-Index

About

Amanda Adkins is a robotics researcher specializing in simultaneous localization and mapping (SLAM) and long-term robot autonomy. Her work addresses one of the field's most pressing challenges: enabling robots to reliably navigate and localize in dynamic, real-world environments that change over time due to geometric shifts, varying viewpoints, and appearance differences. Adkins' most notable contribution, "ObVi-SLAM: Long-Term Object-Visual SLAM" (2024), has garnered 28 citations and represents a significant advance over traditional visual SLAM systems. Rather than relying on fragile low-level feature descriptors that struggle with environmental change and produce unwieldy map sizes, her object-centric approach leverages higher-level semantic representations to achieve more robust and scalable localization. This work builds upon her earlier research, "Probabilistic Object Maps for Long-Term Robot Localization" (2022), which introduced probabilistic frameworks for handling frequent and substantial environmental changes in structured settings like warehouses and parking lots. Together, these contributions position Adkins as an emerging voice in long-term robot deployment research, pushing the field toward more resilient, semantically-aware systems capable of operating consistently across extended time horizons — a critical capability for real-world autonomous robots.