Tommaso Bendinelli Email & Phone Number

Zurich, Switzerland

Alpnach, Obwalden, Switzerland

Leading small teams (2-3 people) to successfully complete projects for our industrial partners, managing priorities, deadlines, budgets, and deliverables. These projects include:- Developing a solution that utilizes a large vision model on construction sites to accurately segment building structural elements- Leveraging millions of unlabeled images via state-of-the-art self-supervised learning techniques- Writing efficient, scalable, and maintainable solution to develop a shared vector database of all internal companydocuments (code, reports) with the goal of simplifying the retrieval process for company knowledg- Using synthetically generated data via Blender for segmentation, pose estimation, and classification of more than 10k different industrial components.- Achieving sub-5mm accuracy in measuring aluminum profiles of 2 meters with only 2D images through key-point estimation and epipolar geometry.

Alpnach, Obwalden, Switzerland

- Developed a computer vision algorithm for the identification of anomalies in the assembly of passive fire protection systems, resulting in a patent application filed by our collaborating industrial partner.- Developed a multi-modal algorithm to automate the recognition of the mining origin of precious gemstones using spectral and chemical data. The solution is currently part of the industrial partner's laboratory's daily analytic routine, resulting in substantial cost savings. - Integrated a stand-alone Python code for smart processing of thermal data into a commercial software via gRPC, enabling our industrial partner to launch the upgraded product in just 3 months.- Collaborated with researchers from ETH Zurich, EPFL, and NASA to produce research papers on the Transformer architecture applied to Science and Engineering, which have been published in leading machine learning conferences and engineering journals.

Alpnach, Obwalden, Switzerland

- Designed a Imitation Learning algorithm for automated robotic teaching which consists of 3D image based 6DOF pose estimation algorithm and automated policy learning algorithm- Hand-eye calibration of a KUKA IIWA robotic system- Visual servoing

Liechtenstein

- Built a Machine Learning pipeline for predictive maintenance of fastening tools.- Contributed in development of an autonomous robot for construction application, specifically: - Developed multiple pieces of software. Examples includes perception and estimation, control, estimation, trajectory generation, and high level task. - Perform extensive Software in the loop, real testing and debugging.- Set up a ROS operating system with various SLAM and SFM packages, interfacing a commercial drone for testing the feasibility of indoor drone accurate surveying.

Zürich Area, Switzerland

I was the project manager of a team of first-year master students during the course, whose scope was to face an industry open challenge. Specifically, my team worked with railway company SBB Cargo on the electrification of rail freights and my responsibility was to support and mentor the students to achieve the specific goals required by the industry partner in the designed time frame

Master'S Degree, Engineering

Bachelor Degree, Engineering