Projects

IIFL: Implicit Interactive Fleet Learning from Heterogeneous Human Supervisors

Conference on Robot Learning (CoRL) 2023

In this work, we propose addressing both multimodality and distribution shift with Implicit Interactive Fleet Learning (IIFL), the first extension of implicit policies to interactive imitation learning (including the single-robot, single-human setting). It achieves 4.5x higher return on human effort in simulation experiments and an 80% higher success rate in a physical block pushing task over (Explicit) IFL, IBC, and other baselines when human supervision is heterogeneous.

Gaurav Datta*, Ryan Hoque*, Anrui Gu, Eugen Solowjow, Ken Goldberg (* equal contribution)

SGTM 2.0: Autonomously Untangling Long Cables using Interactive Perception

IEEE International Conference on Robotics and Automation (ICRA) 2023

This paper extends prior work on autonomously untangling long cables by introducing novel uncertainty quantification metrics and actions that interact with the cable to reduce perception uncertainty. We present Sliding and Grasping for Tangle Manipulation 2.0 (SGTM 2.0), a system that autonomously untangles cables. Experiments suggest that SGTM 2.0 can achieve 83% untangling success on cables with 1 or 2 overhand and figure-8 knots, and 70% termination detection success across these configurations, outperforming SGTM 1.0 by 43% in untangling accuracy and 200% in full rollout speed.

Kaushik Shivakumar*, Vainavi Viswanath*, Anrui Gu, Yahav Avigal, Justin Kerr, Jeffrey Ichnowski, Richard Cheng, Thomas Kollar, Ken Goldberg. (* equal contribution)

PolyPoD: An Algorithm for Polyculture Seed Placement

IEEE International Conference on Automation Science and Engineering (CASE) 2023

We present PolyPoD, a new algorithm for seed placement. Given a polygonal farm boundary, number and types of seeds, PolyPoD generates the variable radius poisson disk distribution to compute viable seed placements. Results suggest that PolyPoD does better than our previous seed placement algorithm in avoiding over-competition, utilization of space, and plant segmentation. Link coming soon.

Varun Kamat, Shrey Aeron, Anrui Gu, Harshika Jalan, Simeon Adebola, Ken Goldberg

CalPERS Divestment Report

Fossil Free California Blog 2021

Analysis of the quantity, value, and trends of CalPERS’s publicly disclosed holdings of equities and bonds in fossil fuel production companies and the broader fossil fuel industry.

Completed during the Undergraduate Research Apprentice Program advised by Professor Clair Brown.

Drone Path Planning

Berkeley FHL Vive Center for Enhanced Reality

Adaptation and analysis of 4D A* and RRT* algorithms for drone obstacle avoidance in simulation.

Covid-19 Predictor

2020 Summer Data Science Exchange

Web app presenting time-series analysis and SIR modeling for the spread of Covid-19 in individual states.