Postdoc: Real-Time AI Vision for Inline Concrete

DTU — Technical University of Denmark; Kongens Lyngby; On-site; DKK 360,000 - 420,000; Full time

Job summary

DTU Construct’s Digital Building Technology section seeks a full-time Postdoc to develop a deep learning–based image analysis framework for in-line monitoring of concrete using hyperspectral imaging. You will enable real-time prediction of material behaviour within production.

You will implement models, train predictive algorithms, and build a GUI for non-experts, collaborating with an interdisciplinary DTU Construct team. PhD required; English fluency expected.

Qualifications

• Background in deep learning, big-data, computer vision, or related fields.
• Background in in-line process monitoring or related fields.
• Preference given to candidates with competence on concrete characterisation and the experimental research.

Responsibilities

• Implement and optimise deep learning–based models for quality control and real-time assessment of concrete constituents within in-line production.
• Develop and train predictive algorithms based on hyperspectral imaging inputs to estimate material properties and performance indicators.
• Develop a GUI to facilitate intuitive application of the model by non-expert users.
• Collaborate with the interdisciplinary DTU Construct and Vistacon teams, especially the other postdoc focusing on material characterisation.
• Disseminate research through publications in top journals and attend conferences.
• Engage in research and teaching activities at the section and departmental level as appropriate.

Skills

Deep learning

Big data

Computer vision

Inline process monitoring

Job description

DTU Construct’s Digital Building Technology section seeks a full-time Postdoc to develop a deep learning–based image analysis framework for in-line monitoring of concrete using hyperspectral imaging. You will enable real-time prediction of material behaviour within production.

You will implement models, train predictive algorithms, and build a GUI for non-experts, collaborating with an interdisciplinary DTU Construct team. PhD required; English fluency expected.