"PhD in AI-driven textile identification for circular economy (1.0 FTE)"

Job Summary

Our project uses sensing technologies such as hyperspectral cameras, NIR and Raman sensors, and an edge-compute AI pipeline to sort used garments into recycle, reuse, and manual-review streams. This PhD project focuses on designing and optimizing a high-throughput hyperspectral imaging system, fused with complementary spectroscopy data and AI, to automatically identify textile fabrics with high accuracy in real-world sorting conditions.

Responsibilities

The PhD candidate will: (1) define optimal spectral bands, spatial resolution, and acquisition speed; (2) develop and validate preprocessing pipelines; (3) architect and compare spectral-only and multimodal deep-learning models; (4) implement robust sensor-fusion strategies under varying conditions; (5) port the inference pipeline to an embedded/edge-compute platform; (6) integrate with a robotic pick-and-place cell for iterative field trials; and (7) disseminate results through peer-reviewed publications and conference presentations while collaborating with the prototyping team.

Qualifications and Requirements

The ideal candidate should have a Master’s degree in computer science or artificial intelligence with a solid grounding in optical and spectroscopic instrumentation. Demonstrated hands-on experience with rapid prototyping, middleware development, and integrating sensing hardware is required. Proficiency in developing AI-driven sensor-fusion pipelines, analytical problem-solving, risk anticipation, and leading interdisciplinary teams is essential. Strong English communication skills are needed.

What the Employer Offers

A four-year fixed-term contract with a salary starting at €3,059 gross per month in the first year, up to €3,881 gross per month in the fourth year, based on a full-time position. Additional benefits include an 8% holiday allowance, an 8.3% year-end bonus, and enrollment in a university PhD training program within the Graduate School of Science and Engineering. The starting date is flexible but must be before 1 November 2025.