I am a PhD candidate in Computer Science and Engineering at the University of Michigan, Ann Arbor. I am advised by Prof. John E. Laird and am a member of the Soar lab.

  preetir (at) umich (dot) edu



My area of research is Human-Robot Interaction; my focus is on learning to characterize a non-expert user's initial mental model of a robot. My goal is to be able to characterize the gaps in the user's mental model, such that this knowledge can be used to bridge the gap between the users' expectations of a robot's capabilities and the robot's actual capabilities, through appropriate feedback mechanisms.


  • August 2019: Attending ACS 2019 (August 2-5) at Massachusetts Institute of Technology in Cambridge, Massachusetts.
  • July 2019: Our paper Towards using transparency mechanisms to build better mental models was accepted to the 7th Goal Reasoning Workshop at ACS 2019!
  • May 2019: I gave a talk at the 39th Soar Workshop presenting results from my internship project.
  • April 2019:Participating in the 2019 CRA-W Grad Cohort Workshop (April 12-13) in Chicago.
  • June 2018: Our extended abstract Establishing Common Ground with Learning Robots was accepted to the RSS 2018 workshop on Models and Representations for Natural Human-Robot Communication!
  • May 2018: I gave a talk at the 38th Soar workshop presenting results from a recent pilot study. (slides)
  • Feb 2018: I will be working as a research intern at Intel Labs, Santa Clara this summer!
  • Oct 2017: My research is once again funded by Intel Corporation for FY 2017!
  • June 2017: I gave a talk at the 37th Soar workshop on my current research project. (slides)
  • July 2016: My research is funded by Intel Corporation for FY 2016.
  • June 2016: I gave a talk at the 36th Soar workshop. (slides)

Past Projects

Mining Insights from Hardware Errata Documents
We use errata documentation to identify interesting patterns and conclusions about product bugs, such as common sources of errors. Over the course of this project, we created a database of over 2,000 different ARM errata and experimented with natural language processing methods ranging from Word2Vec, non-negative matrix factorization and recurrent neural networks. This project was a part of EECS 573 (Microarchitecture) in Fall 2016 implemented as a three person group. Code(GitHub)

Motivated Learning – Replication project
This is a replicated project based on the study specified in “Graham, J., Starzyk, J. A., Ni, Z., He, H., Teng, T. H., & Tan, A. H. (2015, July). A comparative study between motivated learning and reinforcement learning. In 2015 International Joint Conference on Neural Networks (IJCNN)(pp. 1-8). IEEE.” in order to test the hypothesis that Motivated Learning earns a higher average reward than Reinforcement Learning in the custom dynamic environment specified. The hypothesis was replicated successfully. This project was a part of EECS 592 (Advanced Artificial Intelligence) in Winter 2016.

Index-based Load Optimization in MySQL
Implementation of an automated drop-and-rebuild-indexes scheme around a load operation to improve performance of loading huge data into databases in MySQL. This project was a part of EECS 584 (Advanced Database systems) in Fall 2015 implemented as a two person group. Code(GitHub)