Cornell researchers studying microplastics, robotics and machine learning are recent recipients of National Science Foundation Faculty Early Career Development Awards.

Grantees receive a minimum of $400,000 over five years from the program, which supports early-career faculty who “have the potential to serve as academic role models in research and education and to lead advances in the mission of their department or organization,” according to the NSF. Each project must include an educational component.

Natalie L. Cápiro ’00, assistant professor of biological and environmental engineering in the College of Agriculture and Life Sciences, will use her award to study how microplastics move through soil and water and interact with potentially toxic halogenated organic compounds (HOCs) such as chlorinated ethenes and per- and polyfluoroalkyl substances (PFAS).

The research will explore the environmental fate and transport of microplastics and their interactions with associated adsorbed HOCs through a series of laboratory experiments that increase in scale and complexity, and are designed to mimic conditions at contaminated sites. Objectives include evaluating the impacts of aged and weathered microplastics on HOC adsorption and biotransformation kinetics, determining the effects of flow (e.g., simulated rainfall events) on the fate and transport of microplastics and adsorbed HOCs, and creating programs to disseminate knowledge about the impact of HOCs and microplastics on water quality. Results of the project will advance scientific knowledge that can be used to protect public health, remediate contaminated sites and protect water supplies.

Sanjiban Choudhury, assistant professor of computer science in the Cornell Ann S. Bowers College of Computing and Information Science, will use his award to support his work creating robots that learn new skills the way humans do. Robots can provide assistance at home, in hospitals and on farms, but most robots can only do tasks that are pre-programmed ahead of time. They cannot handle new situations or learn from people. This project supports research to create robot helpers that learn new skills by watching people, trying tasks and improving from feedback.

This work could make robots more helpful and flexible, in order to solve harder problems in the real world. The work also intends to help us understand how robots can learn and adapt. The project looks to improve robots for homes, health care and agriculture, and includes educational programs with interactive robotics activities for K-12 students. It provides accessible online resources to increase participation in STEM and robotics research. 

Sarah Dean, assistant professor of computer science (Cornell Bowers), will use her award to develop algorithms for reliable decision-making in complex systems, ensuring that the benefits of machine learning are realized while minimizing its risks. Feedback loops – when information about system output is used in system input – are crucial because they affect how models learn from data and interact with their environment.

Understanding feedback can help machine learning improve outcomes in areas like weather prediction, recommendation systems and human-robot collaboration. Dean's research program will develop theory and algorithms that have direct impact on applications in weather prediction, autonomous aerial navigation, recommendation systems, and human-robot interaction.  In addition, Dean will develop hands-on projects to introduce high school students to exciting topics like AI weather forecasting, balloon control and robotics.