Chadi Salmi

Chadi Salmi

AI & Robotics Engineer passionate about intelligent systems

2024 - Present
TU Delft

Robotics Software Architect, Viscon

Design the data collection and processing pipeline for an autonomous greenhouse robot that can navigate, collect data, and analyze the plant.

2021 - 2024
TU Delft

AI and Robotics Team Lead, AIRLab

Led the development of a robotic platform for autonomous grocery item picking, applying advanced computer vision techniques for 6 DoF pose detection and item classification. Pioneered the integration of few-shot learning models, enabling adaptive product classification without the need for retraining per item.

2021
TU Delft

Research Engineer, TU Delft

Developed and implemented a vision-based pedestrian detector for 3D detection and tracking on a mobile robot platform, leveraging both LiDAR and camera data.

2020 - 2021
TU Delft

MSc Thesis - Continual Learning for Pedestrian Motion Prediction

Investigated state-of-the-art continual learning techniques for enhancing pedestrian motion prediction models in dynamic scenarios. Designed and deployed a self-supervised learning framework on an autonomous robot platform.

2018 - 2019
TU Delft

Deep Learning Engineer, Formula Student Delft

Developed and implemented computer vision algorithms for lane detection and object recognition, advancing the team's capabilities in autonomous navigation and decision-making.

2016 - 2020
TU Delft

Full-stack Developer (part-time)

Developed features for a restaurant POS web application, utilizing modern web technologies and frameworks.

About Me

Experienced and forward‑thinking robotics professional with 5+ years of hands‑on experience developing advanced manipulation systems. Proficient in software architecture applied to robotics, computer vision, and machine learning, with a passion for tackling complex engineering challenges and delivering reliable, high‑impact solutions.

Education

M.S. in Cognitive Robotics

Technical University of Delft

September 2017 - March 2021

B.Eng. in Mechanical Engineering

Technical University of Delft

September 2014 - July 2017

Skills

Programming Languages

Python C++ Rust JavaScript TypeScript HTML CSS C#

Tools & Frameworks

Git PyTorch ROS Linux React Dotnet

Languages

Dutch Native
English Fluent

Publications

Improving Pedestrian Prediction Models With Self-Supervised Continual Learning

Chadi Salmi, Luzia Knoedler, Hai Zhu, Bruno Brito, Javier Alonso-Mora

IEEE Robotics and Automation Letters (RA-L)

2022

Sampling-based Model Predictive Control Leveraging Parallelizable Physics Simulations

Corrado Pezzato, Chadi Salmi, Max Spahn, Elia Trevisan, Javier Alonso-Mora, Carlos Hernandez Corbato

IEEE Robotics and Automation Letters (RA-L)

Earlier version in ICRA Workshop "Embracing Contacts", 2023 — Received 3rd Best Paper

2025

Local Planner Bench: Benchmarking for Local Motion Planning

Max Spahn, Chadi Salmi, Javier Alonso-Mora

IROS Workshop "Evaluating Motion Planning Performance"

2022