Viktor Makoviychuk
Papers
13
Total Citations
741
H-Index
9
About
Viktor Makoviychuk is a pioneering researcher at the intersection of physics simulation, GPU computing, and robot learning, whose work has fundamentally shaped how modern reinforcement learning is applied to complex robotic manipulation. He is perhaps best known as a lead contributor to **Isaac Gym** (2021, 322 citations), NVIDIA's landmark GPU-accelerated physics simulation platform that revolutionized robot learning by enabling massively parallel training directly on GPU hardware — eliminating costly CPU-GPU data transfers and dramatically accelerating policy development. This work built on his earlier GPU-accelerated simulation research dating to 2018, establishing a coherent vision for scalable robot learning infrastructure. Makoviychuk has made equally significant contributions to **dexterous in-hand manipulation**, tackling the notorious sim-to-real transfer problem through projects like DeXtreme (88 citations) and multiple TriFinger manipulation studies. His research on adaptive simulation randomization further addresses the sim-to-real gap by intelligently tuning simulation parameters using real-world experience. Beyond robotics, his work on non-smooth Newton methods for deformable multi-body dynamics (83 citations) demonstrates deep expertise in computational physics. Collectively, his publications have accumulated over 700 citations, making him one of the most influential figures in GPU-driven robot learning today.
Research Focus
Key Achievements
Top Papers
- 1Isaac Gym: High Performance GPU-Based Physics Simulation For Robot\n Learning322 citations · 2021
- 2DeXtreme: Transfer of Agile In-hand Manipulation from Simulation to Reality88 citations · 2023
- 3Non-smooth Newton Methods for Deformable Multi-body Dynamics83 citations · 2019
- 4GPU-Accelerated Robotic Simulation for Distributed Reinforcement\n Learning74 citations · 2018
- 5
- 6
- 7
- 8
- 9
- 10