Darshak Panseriya

Perception Engineer

Camera-radar fusion and tracking on public roadside recordings (INFRA-3DRC): boxes matched to radar carry its range and radial speed, and each road user keeps a track ID.Code on GitHub

I build and validate ML perception on real sensors, with a background in suspension design and vehicle dynamics.

At Bosch I took a radar kick-detection function into series development in under seven months: data campaigns, the model, the detection system and a live demonstrator, then robustness in a customer vehicle and a passed OEM design review. Its architecture is filed as an internal invention disclosure.

Stuttgart · German C1 (telc) · English C1

See the perception work About and CV

Perception & ML

Radar sensing, sensor fusion and machine learning on real automotive hardware.

Robert Bosch GmbH · Oct 2025 to Sep 2026

Radar kick detection at Bosch

Hands-free tailgate opening, sensed by the digital-key radar already in the car. The tools behind it were adopted by Bosch teams in three countries.

RAM budget for the whole function, set by the microcontroller target: ~100 KB

Open the project

Drawn to explain; not Bosch data.

A road at night from a roadside camera and radar: two cars, each in a labelled box with its track number, range and radial speed, one approaching and one driving away; beside it, a bird's-eye view of both tracks with their history lines and velocity arrows — Fusion and tracking at night, INFRA-3DRC scene 25SourceSensor-Data-Fusion, 2026 · INFRA-3DRC by Fraunhofer IVI, CC BY-NC 4.0

Fraunhofer IVI · Mar 2024 to Jun 2024

Camera-radar fusion and tracking on real roadside recordings

I led a student team of four and the technical work; in 2026 I added a tracker and scored it over 22 scenes.

Open the project Code on GitHub

One cone three times: the camera image, the outline traced from its FSOCO mask, and the predicted keypoints scored by their distance to that outline — One cone's keypoints, scored against its mask outlineSourceCone-Pose-Validation, FSOCO images

Schanzer Racing Electric, Formula Student Driverless · Oct 2024 to Apr 2025

A keypoint detector validated from existing segmentation masks, with no new labels

Mask outlines as ground truth, on images from more than 40 teams' cars and cameras.

Open the project Code on GitHub

UWB radar evaluation and toolingRobert Bosch GmbH · Internship · Feb 2025 to Sep 20251-2 h → secondssignal analysis per measurement

All perception work

Vehicle Design & Dynamics

Where I started: suspension design and vehicle dynamics, on a micromobility platform and a Formula Student race car.

Three-person micromobility startup · Aug 2022 to Aug 2024

A four-wheel micromobility vehicle under 50 kg, from vision to prototype design

Sole vehicle platform design engineer: from the founder's vision to a concept, and from the concept to a manufacturable prototype design, covering sketches, part design, assembly and the bill of materials.

GTU Motorsports · Apr 2019 to Mar 2021

Vehicle dynamics for a Formula Student car, 3rd at Formula Bharat 2020

cut from already-light suspension parts: ~3 kg

All vehicle design and dynamics work

Builder Lab. Tools I build to run my own work, with AI coding agents: DeckHQ (released on npm) and MechCopilot (open source).

Open the Builder Lab

About

I hold an M.Eng. in Automated Driving and Vehicle Safety from TH Ingolstadt and a B.Tech in Automobile Engineering from Parul University.

Full experience and CV