Gijs van den Brandt

PhD Candidate Robotics

Eindhoven, North Brabant, Netherlands

About

Experience

  • Doctoral Candidate at Eindhoven University of Technology
    Jul 2023 - Present · 3 yrs

    Supervisors: René van de Molengraft, Elena Torta and Jordy Senden The goal of my research is to increase robot autonomy by modeling invariant properties of the world. To give an example, rather than hard-coding movements that result in a specific robot manipulating a specific object, we model properties such as the object geometry so that the robot may autonomously derive which actions it should take to complete a task. An open question is how to deal with uncertainty of the modeled properties. World models benefit, e.g., manufacturing SMEs with small production volumes. Currently, these companies rely on manual labor to maintain agile production lines in an unstructured environment. Programming a robot to perform these tasks robustly requires an expensive expert who must revise the program every time the product changes. World models can reduce programming costs and make robots a cost-effective alternative, even for small companies.

  • Graduate Student Researcher at I.AM. H2020 Project
    May 2022 - Mar 2023 · 11 mos

    Supervisors: Alessandro Saccon, Jens Kober and Jari van Steen My Msc. thesis was conducted as part of the I.AM. project. My goal was to experimentally validate reference spreading -- a control method that allows robots to exploit impacts, e.g., during object manipulation. This is interesting for industry as it can speed up logistic tasks, such as removing boxes from a pallet. Evaluating reference spreading for box grabbing on a dual-arm robot setup involved: designing a soft end effector, implementing impedance control, planning references with teleoperation, and expanding on the reference spreading framework to prevent reference discontinuities and handle delayed impact detection.

  • Teaching Assistant at EuFlex Technificent
    Feb 2020 - Jan 2023 · 3 yrs

    While working at EuFlex, I guided classes of 10-50 TU/e students working on homework assignments. These classes were either on-campus or online. I was a TA for the following courses: - Solid Mechanics (2020) - Intro to Mechanical Engineering (2020) - Dynamics (2020, 2021, 2022) - Optimal Control & Reinforcement Learning (2021)

  • Intern Innovate Department at Vanderlande
    Sep 2021 - Nov 2021 · 3 mos

    My internship at Vanderlande concerned the detection of objects from LiDAR measurements using machine learning. I worked independently to setup a simu- lation environment for synthetic data generation, and subsequently designed a convolutional neural network.

  • Machine Operator at Broflow B.V.
    Jul 2015 - Oct 2020 · 5 yrs 4 mos

    At Broflow I worked on packaging vegetables and fruit for grocery stores. This included checking the quality, packaging in boxes, and labelling. My promotion to Machine Operator added the responsibilities of supervising a small team, configuring labelling machines, and operating flow-packing machines