Max Robotics — Global Service Robot Directory + US Certification

About

Nikhil Mishra advances the frontier of embodied intelligence, bridging model-based control, perception under uncertainty, and sim-to-real transfer for autonomous systems. His foundational work on combining model-based policy search with online model learning enabled physical humanoid robots to autonomously acquire control policies for high-level tasks without manual trajectory specification—a contribution that has garnered 53 citations and laid groundwork for adaptive robot learning. Mishra further explored socially-aware robotics through an ego-centric framework for telepresence robots, addressing the challenge of natural human-robot interaction in shared spaces. Recognizing the critical need for reliability in autonomous perception, he developed autoregressive uncertainty modeling for 3D bounding box prediction and introduced Distributional Instance Segmentation (Latent-MaskRCNN), which captures meaningful uncertainty in ambiguous scenes to prevent catastrophic errors in high-stakes applications. Most recently, Mishra’s work on closing the visual sim-to-real gap using Object-Composable NeRFs addresses the persistent challenge of costly real-world data collection, enabling more effective domain randomization for perception systems. His research trajectory—from control to uncertainty-aware perception to simulation fidelity—demonstrates a systematic approach to building robust, deployable robotic intelligence.