Skip to main content

News

Amanda Adkins

Amanda Adkins is a Ph.D. student in the Texas Robotics Graduate Portfolio Program at the University of Texas at Austin. Her work is focused on perception for long-term autonomy, specifically at the intersection of Simultaneous Localization and Mapping. Through her work, she is hoping to make real systems much more accurate, robust, and easier to be deployed by non-experts in real environments. Her professional goal is to work in industry or for the government on mobile robots or autonomous cars. 

Amanda Adkins

Amanda Adkins is a Ph.D. student at the University of Texas focusing on robotics. She works in the Autonomous Mobile Robotics Lab with Professor Joydeep Biswas. After getting her undergraduate degree at Worcester Polytechnic Institute with a double major in Computer Science and Robotic Engineering, she went to work for Amazon Robotics as a software developer. There, she completed work highlighting problems within localization and mapping. She was drawn to UT Austin for her Ph.D. because of the increasing focus on robotics at UT and her advisor’s work. 

Currently, Amanda works on perception for long-term autonomy. She specifically focuses on the intersection of Simultaneous Localization and Mapping (SLAM) and semantic scene understanding for long-term deployments in robots. With this research comes challenges, such as scaling, lighting changes, and moving elements that Amanda must face and overcome. While these challenges increase the difficulty of the work, Amanda finds addressing them valuable because long-term deployments enable robots to improve their perception over time and learn from their experiences. 

Amanda aims to graduate in 2025 and find work in industry or government research on problems for mobile robots or autonomous cars. She would like to spend time researching and publishing papers along with applied work in which she can relate her ideas to specific problems. Having hands-on experience is something that Amanda has valued the most from her time in research. With the hands-on research she has accomplished, she has identified problems that can make a real impact in robotics and the real world. 

Before graduating, Amanda has a bit more work she wants to accomplish. She recently submitted her second paper on long-term object-visual SLAM and has just begun working on a new project on object association for semantic SLAM. She hopes to complete two or three more projects before graduating that will extend her work to incorporate more machine-learning techniques and better object identification and association. 

The potential impact of Amanda’s work is outstanding. Her research ideas and code development could make real systems more accurate, robust, and accessible for non-experts to deploy in real environments. This will hopefully extend to delivery robots, warehouses, and homes. 

Her work displays the dedication and ability of students in Texas Robotics. They strive for innovation and success by moving society forward and integrating technology in many ways.

 

During her research, Amanda has worked with Taijing Chen, Joydeep Biswas, and Shakeel Samsudeen. She is currently collaborating with Dongmyeong Lee.