ROBOTICS R&D ENGINEER
Lakshya Jain
MS Robotics & Autonomous Systems, ASU. Building causal RL systems, embodied AI, and real-world robot deployments.
CAUSAL RL
ROS2 / MOVEIT
EMBODIED AI
SIM-TO-REAL
FEATURED PROJECTS
ASU THESIS · NOV 2025
Discover-Intervene-Adapt (DIA)
Interpretable causal RL with structured world models and intervention strategies. 35%
reduction in Montezuma exploration time, 15x CoinRun generalization improvement.
UNREAL ENGINE 5.4 · ASU
AI Therapist in VR
AI-powered conversational agent for mental health therapy in VR. RAG LLM based on LLAMA with speech recognition and adaptive RL dialogue generation.
1ST PLACE
DEVIL'S INVENT HACKATHON · DEC 2024
IntuitionAI
RAG-based adaptive tutoring agent using multi-modal transformers and RL for
curriculum sequencing. 40% perceived learning efficiency gain.
LOGOS LAB, ASU · IROS 2026 (SUBMITTED)
Meanings & Measurements: Multi-Agent Probabilistic Grounding (MAPG)
Multi-agent probabilistic framework for vision-language navigation. Grounds
natural-language instructions to metric-semantic scene representations, surpassing VLM baselines by 20%
accuracy.
THREE.JS · PERSONAL PROJECT
K-VRC
Real-time 3D robot avatar with a generative emotion and facial expression system trained on demonstration data from Netflix's Love, Death + Robots. Emotion state drives procedural lighting, bloom, and body language. Chat interface runs a live LLM with voice input; a sidenote panel surfaces contextual project insights as you explore the portfolio.
GOVT. FUNDED
LTA LAB, IIT BOMBAY · DST
Project HERCARA
Govt. funded (Dept. of Science & Tech.) swarm of VTOL autonomous UAVs, aerostat,
robotic arms, and embedded safety systems.
PERSONAL PROJECT · OCT 2025
RobotChessPlayer
Two Franka Panda arms on opposite sides of a shared board in Isaac Sim 4.2 perceive game state, plan staged motions with MoveIt 2, and execute alternating turns via a ROS 2 coordinator. White plays classical engine; black runs a learned RL policy behind the same FEN-in / UCI-out interface.
3RD PLACE
HONEYWELL ANTHEM HACKATHON
Flight Autopilot System
RAG LLM-based autopilot taking over all autopilot tasks and automating plane
features. Top-3 at Honeywell's Anthem Hackathon for aviation AI.
#6 WORLDWIDE
SAE AERO DESIGN EAST · 2022
Rayar RC Airplane
Advanced class RC airplane capable of payload drop and small glider deployment.
Gliders autonomously fly a custom sequence. #6 worldwide.
PUBLISHED
KJSCE GRADUATE PROJECT
Forest Surveillance Rover
Swarm of RPi & Arduino Mega rovers with object, animal, and fire detection,
autonomous navigation, and multi-modal operation. Published in IJRASET.
ASU PERSONAL PROJECT · AUG 2025
Snydrone
LLM-driven ROS2 cinematic drone planner translating natural-language prompts to
trajectories. 30% framing consistency improvement in simulation.
EXPERIENCE
AUG 2024 — DEC 2025
Graduate Teaching Assistant
Arizona State University
Delivered lectures for Perception in Robotics and Data Structures & Algorithms.
Improved student assessment performance by 20% through structured lesson design. Mentored projects on
transformers, attention, and multi-modal LLM pipelines.
AUG 2024 — NOV 2025
ML Research Aide
DeSmart Lab, NSF Brain Center
Built adaptive RL-driven lighting prototypes for medical applications. Achieved 15%
below-SOTA NASA TLX mental demand ratings. Implemented end-to-end ML pipeline with automated tuning and
AWS/GCP deployment.
AUG 2022 — JUL 2023
R&D Systems Engineer
Indrones
Developed autonomous drone ML pipelines for surveillance and target tracking. Enabled
reliable GPS-denied operations at 6000+ m Himalayan altitudes. Built predictive maintenance toolkit
reducing system failures by 50%.
JUN 2021 — MAR 2022
Research Intern
LTA Lab, IIT Bombay
Created deep-RL precision landing system for VTOL aircraft achieving 2 cm visual test
accuracy. Implemented edge ML weather and emergency recognition at 99% accuracy.
TEACHING — CSE 310 DATA STRUCTURES & ALGORITHMS
Bubble & Insertion Sort
VIEW SLIDES →
Asymptotic Analysis
VIEW SLIDES →
Recurrence Relations
VIEW SLIDES →
Heaps, Heapsort & Priority Queues
VIEW SLIDES →
BST / DFS / BFS
VIEW SLIDES →
Greedy Method
VIEW SLIDES →
Dijkstra's Algorithm
VIEW SLIDES →
Red-Black Trees
VIEW SLIDES →
Hash Tables
VIEW SLIDES →
Project 2 Walkthrough
VIEW SLIDES →
Exam Review
VIEW SLIDES →
Exam Review: DFS, Edge Classification & Topological Sort
VIEW SLIDES →
Final Exam Review
VIEW SLIDES →
PUBLICATIONS & DISTINCTIONS
PUBLICATIONS
- IROS 2026 (Submitted): Meanings and Measurements: Multi-Agent Probabilistic Grounding for Vision Language Navigation — Project Site →
- NeurIPS 2026 (In Progress): Discover, Intervene, Adapt: Causal RL with Interleaved Discovery and Learning
- IJRASET: Multi-level architecture for a swarm of surveillance rovers — DOI →
- IJRASET: Image Classification and Object Following Functions for Mobile Robots — DOI →
ACHIEVEMENTS
- Impact Award Nominee, ASU
- 1st Place — Devil's Invent Hackathon
- 3rd Place — Honeywell Anthem Hackathon
- #6 Worldwide — SAE Aero Design East Advanced Class 2022
- #2 Worldwide — SAE ADE Regular Design Report 2020
- TSMC AZ Fellowship · Ira A. Fulton Fellowship · ASU Grad Fellowship
TECHNICAL SKILLS
CAUSAL RL
CAUSAL DISCOVERY
DEEP RL
CURRICULUM LEARNING
IMITATION LEARNING
WORLD MODELS
REWARD SHAPING
INTRINSIC MOTIVATION
ROS / ROS2
MOVEIT
SLAM
NAV2
MPC
PATH PLANNING (A*, RRT)
SENSOR FUSION
URDF / XACRO
TRAJECTORY CONTROL
ARDUPILOT / PX4
ROS2 CONTROL
RAG SYSTEMS
TRANSFORMERS
NLP
MULTI-MODAL LLMs
LANGCHAIN
LLAMA
PROMPT ENGINEERING
SCENE GRAPHS
OBJECT DETECTION
POINT CLOUD (PCL)
DEPTH ESTIMATION
YOLO
VISUAL NAVIGATION
PYTORCH
TENSORFLOW
DEEP LEARNING
PROBABILISTIC INFERENCE
GRAPH NEURAL NETWORKS
ISAAC SIM
GAZEBO
PYBULLET
ISAAC LAB
UNREAL ENGINE 5
SOLIDWORKS
FUSION 360
CAD MODELLING
ROBOT KINEMATICS
PYTHON
C++
RASPBERRY PI
ARDUINO
EMBEDDED LINUX
DOCKER
KUBERNETES
AWS / GCP
WANDB / MLFLOW
GIT / CI-CD