Vainavi Viswanath
Vainavi Viswanath researches robot perception and manipulation of cables. She began undergraduate research at Berkeley’s Autolab in February 2020, working on perception-driven untangling. She coauthored the 2022 system for autonomously untangling long cables and the 2023 HANDLOOM paper on learned tracing of one-dimensional objects. The latter combines neural-network tracing with classification of crossings and robot manipulation. Its experiments concern defined cable configurations and materials. Her work connects visual perception with physical actions.
- Works
- 9
- Citations
- 71
- h-index
- 5
Publication numbers from OpenAlex, as of 14 October 2026.
Timeline
-
February 2020
Autolab research
Viswanath began undergraduate research at Ken Goldberg’s Berkeley Autolab in February 2020. She worked on perception-driven untangling of ropes.
-
2022
Untangling long cables
She coauthored the RSS system on autonomous untangling of long cables. The system uses robot motion primitives, specialized grippers and cable-slack management. Its tests involve specified knots and crossings, with success limited to the reported configurations.
-
2023
Learning to trace cables
Viswanath and Kaushik Shivakumar coauthored HANDLOOM with equal contribution, alongside a wider author team. It learns cable traces and crossing classes from simulated and real examples, using greyscale images. The paper’s scope is semi-planar configurations.
-
2023
From tracing to manipulation
The system combines learned tracing with a two-arm robot for knot tying and untangling. Tests covered defined materials and knot configurations.The system links visual state estimates with robot actions.
Sources and credits
Photo: web.archive.org; People | AUTOLab; creator not stated. License: Not stated by source. Source: web.archive.org. Cropped to a square and resized. Ask us to remove or replace this photo.