Open to opportunities

Saran Sundar
Subbiah

Robotics Engineer /

MSc Robotics · TU Delft  ·  Robotics Engineer · RoboHouse
Building autonomous systems that work in the real world.

View Projects Get in Touch Download CV
scroll
01 / About
Saran Sundar Subbiah
Delft, Netherlands

Building robots that
understand the world.

I'm a Robotics Engineer currently pursuing my MSc at TU Delft one of Europe's top engineering universities. My work sits at the intersection of motion planning, autonomous navigation, and real-world deployment.

Alongside my studies, I work part-time at RoboHouse as a Software Engineer, where I built a skid-steer robot from the ground up, from hardware integration to the full ROS2 autonomy stack. That platform is now the foundation of my thesis: a terrain-aware MPPI controller I'm developing under the project I call SLIP Scalable Locomotion for Imperfect Places.

Before Delft, I spent my time in industrial robotics and automation working with Dassault Systèmes, with startups on AMRs, and competing in hackathons. I care about robots that actually work outside the lab.

MSc Robotics Delft University of Technology
2024 – Present · Delft, NL
BE Mechatronics Engineering Sathyabama IST
2019 – 2023 · Chennai, IN
5+ Competition Podiums
02 / Experience

Where I've worked.

Aug 2024 — Present
Delft, NL

Robotics Software Engineer

RoboHouse · TU Delft

Building a skid-steer robot from scratch — mechanical integration, ROS2 software stack, and autonomous navigation. The platform is the foundation for my MSc thesis on terrain-aware motion planning (Project SLIP).

ROS2 C++ Python MPPI Nav2 SLAM
Oct 2023 — Aug 2024
Chennai, IN

Junior Robotics Engineer

Doozy Robotics

Designed, built, and tested robotic systems with emphasis on ROS. Worked on AMRs using SLAM for autonomous navigation and mapping, and contributed to early-stage humanoid robot development.

ROS SLAM AMR Python
Oct 2022 — Jun 2023
Chennai, IN

Industrial Technical Consultant

Dassault Systèmes

Led robotics programming, simulation, and virtual commissioning using DELMIA and CATIA for industrial clients. Optimized robotic automation workflows from concept to virtual deployment.

DELMIA CATIA 3DEXPERIENCE Virtual Commissioning
Oct 2021 — May 2022
Chennai, IN

Automation Engineer

HiMAC

Studied and reverse-engineered Special Purpose Machines. Programmed PLC control panels, produced detailed design documentation, and built test rigs for complex electromechanical systems.

PLC SCADA Mechanical Design Siemens NX
Oct 2020 — May 2021
Chennai, IN

Student Intern

Micro Designs

Supervised production projects and optimized mechanical component design workflows for industrial applications. Worked hands-on with CNC machinery — VMC, CNC milling, and die engraving.

CNC Machining VMC CAD Manufacturing
03 / Projects

Things I've built.

SLIP robot 1 / 6
MSc Thesis · RoboHouse · In Progress

SLIP — Scalable Locomotion for Imperfect Places

Built a full skid-steer robot from scratch at RoboHouse — mechanical assembly, electronics integration, and a complete ROS2 autonomy stack. The platform is now the foundation of my MSc thesis: a terrain-aware MPPI controller that adapts motion planning in real-time based on terrain classification, enabling reliable outdoor navigation where standard planners fail.

MPPI Terrain Estimation ROS2 C++ PyTorch SLAM Embedded Systems Stochastic Control
GitHub
1 / 4
Machine Learning for Robotics

Intelligent Control Systems for Robotic Manipulators

A suite of intelligent control and ML techniques applied to robotic arm control — CNN-based joint angle estimation from images, physics-informed Lagrangian Neural Networks (LNN) in JAX/Flax for model-based control, Iterative Learning Control (PD-ILC & Q-ILC) for precise trajectory tracking, and Gaussian Process Regression for dynamics learning and uncertainty-aware policy cloning from demonstrations.

Python JAX Flax PyTorch LNN ILC Gaussian Processes
Techfest IIT Bombay · 2nd Place · ₹40,000

Find the LAG — Ubisoft

Frame-by-frame structural similarity analysis (SSIM) to detect input lag in games — outperformed traditional background subtraction methods in both accuracy and speed.

Computer Vision SSIM Python
GitHub
ITU AI for Good · TinyML · 2nd Place

Plant Disease Classification

Custom TinyML architecture deployed on XIAO ESP32-S3 for on-device plant disease detection. Achieved 74% accuracy with minimal parameters under strict edge deployment constraints.

TinyML ESP32 Edge AI Python
GitHub
Chennai Cybersecurity Hackathon · 3rd Place

CCTV Footage Enhancement

ResAutoencoder pipeline to enhance pixelated CCTV footage trained with synthetic degradation. Achieved 98.6% SSIM accuracy — enabling usable forensic footage from low-quality inputs.

Deep Learning Autoencoder Computer Vision GCP
GitHub
Multilingual Semantic Search Hackathon

R PaperPal

AI engine for researchers — recommends semantically similar papers and translates them to the user's native language using MBART Large-50 and Cohere summarization.

NLP MBART Cohere Semantic Search
View
04 / Skills

Tools & technologies.

Robotics & Autonomy

ROS / ROS2core
SLAMcore
MPPIcore
Nav2core
Motion Planningcore
Robot Dynamics & Controlproficient

Programming

Pythoncore
C++core
Cproficient
MATLABproficient
Bash / Linuxproficient

AI & ML

PyTorchcore
Computer Visioncore
OpenCVproficient
TensorFlow / TinyMLproficient
Edge AIfamiliar

CAD & Simulation

Gazebocore
Siemens NXproficient
CATIA / DELMIAproficient
Fusion 360familiar
AutoCADfamiliar

Industrial

PLC Programmingproficient
Virtual Commissioningproficient
SCADAfamiliar
3DEXPERIENCEfamiliar
05 / Recognition

Awards & publications.

2nd

Ubisoft — Find the LAG

Techfest, IIT Bombay  ·  ₹40,000 prize

2nd

ITU TinyML — Plant Disease Detection

AI for Good · International Telecommunication Union

3rd

ITU TinyML — Wildlife Monitoring

AI for Good · International Telecommunication Union

3rd

Chennai Cybersecurity Hackathon

CCTV Footage Enhancement  ·  98.6% SSIM accuracy

★

Star of Excellence Award

Mechatronics Batch 2023 · Sathyabama Institute of Science & Technology

06 / Contact

Let's talk.

I'm always open to discussing robotics research, collaborations, or full-time opportunities. Drop me a message — I'll get back to you.

saransrimathi590@gmail.com
LinkedIn · GitHub · Instagram · Reddit