Contractual lecturer-researcher AI for Industry

Teaching

The successful candidate will teach 64 hours per year to strengthen educational activities in the areas of data science, artificial intelligence, and intelligent interactive systems (e.g., machine learning, trustworthy AI, AI-driven decision-making, human-centered AI, data-driven systems, intelligent perception, explainable AI).

The recruited faculty member will contribute to existing courses and may develop new ones. Teaching activities will mainly target Master-level programs, in synergy with the new ERASMUS MUNDUS program European Master in Sustainable Systems Engineering (UTC, University of Genoa – Italy, Universitat Politècnica de Catalunya – Spain, and Polytechnic University of Tirana – Albania).

The successful candidate will also have the opportunity to contribute to the UTC’s computer science engineering programme for students undergoing initial training (FISE) and alternating training with companies (FISA).

Research

This position is part of the "AI for Industry and Risk Management" CAP (Collaborative Acceleration Program) initiative, from the PostGenAI@Paris hub coordinated by Sorbonne Université.

Objectives:

The project addresses a central challenge in modern industrial AI: moving from high-performing predictive systems toward trustworthy, auditable AI systems suitable to risk analysis for deployment in critical operational environments.

While current AI technologies have demonstrated remarkable performance in areas such as computer vision, predictive maintenance, anomaly detection, quality inspection, forecasting, and decision support, their adoption in industrial contexts remains limited due to several structural limitations:

·        limited interpretability of model behavior;

·        weak guarantees regarding robustness and reliability;

·        insufficient characterization of predictive uncertainty;

·        lack of robustness to distribution shifts, rare events, and adversarial conditions;

·        lack of traceability and auditability required by industrial and regulatory standards.

The project therefore aims to establish a new methodological framework for Trustworthy Industrial AI, combining statistical rigor, explainability of predictions, uncertainty modeling, compliant with a human-centered supervision.

The selected candidate will join the Heudiasyc laboratory and contribute to one or more of the project’s scientific axes. The goal is not merely a purely technical vision of prediction tools, but also regulatory compliance and operational trust (industrial acceptance, improved human oversight).

·        From raw prediction to reliable decision support

o   estimating confidence levels associated with predictions;

o   identifying situations outside their operational validity domain;

o   supporting risk-sensitive industrial decision-making and communicating uncertainty to human operator.

·        Explainability and interpretability of AI systems

o   Development of AI models whose behavior can be analyzed, interpreted and audited by domain experts;

o   explainable machine learning, post-hoc explanation techniques;

o   attribution methods;

o   causal and symbolic reasoning approaches;

o   human-understandable representations of AI decisions.

·        Quantification and Propagation of uncertainty in industrial environments (noisy sensors, sensor degradation, evolving production processes, rare events, incomplete datasets…)

o   quantifying epistemic and aleatoric uncertainty;

o   detecting abnormal or unforeseen situations, domain shift;

o   calibrating confidence estimates;

o   propagating uncertainty through AI pipelines.

These research directions may be taken using different approaches such as Bayesian learning, evidential learning or conformal prediction to provide AI systems with mechanisms allowing them to “know when they do not know", provide explanations, trigger human intervention when necessary, and reduce unsafe autonomous decisions. The program seeks to improve robustness with respect to several effects, eventually leading towards safety guarantees for machine learning systems, validation protocols and certification frameworks. A long-term ambition is to contribute to future standards and methodologies for the certification of trustworthy industrial AI systems. 

Research activities may be validated using simulation platforms or experimental systems or data.

The recruited candidate will also contribute to the project’s partnerships and development strategy. The program is inherently interdisciplinary, combining research in artificial intelligence, machine learning, human-computer interaction and interactive systems.

ADDITIONAL INFORMATION

Additional activities

Participate in the implementation of UTC’s Sustainable Development & Social and Environmental Responsability (SD&SER) Master plan.

Organisation

Université de technologie de Compiègne (UTC), a member of the Sorbonne University Alliance (ASU) and the network of universities of technology (UT), is ranked among the top French engineering schools in a number of national rankings, and offers a particularly favorable environment for teaching and research.

https://www.utc.fr/en/

The department

The Department of Computer Science, one of the six departments at UTC, offers course units for entry-level students as part of the UTC Common Core, as well as for students pursuing different engineering majors (whether full-time or as a sandwich course). It also provides professional vocational training in engineering. Additionally, the department also awards research degrees at the master’s and PhD levels.

The Department hosts the LMAC and Heudiasyc laboratories. It maintains strong connections with industry in both teaching and research, and has established close links with international academic institutions and partners.

https://www.utc.fr/en/courses-and-training/the-utc-engineering-diploma/computer-sciences-and-engineering-gi/

The laboratory

Heudiasyc (UMR 7253) is a joint research unit supported by UTC and CNRS. It conducts multidisciplinary research focused on information science and technologies, including artificial intelligence, machine learning, uncertain reasoning, operational research, networks, robotics, automation, and knowledge representation.

Heudiasyc’s activities are based on a synergy between basic research and technological research to better address the major societal challenges in the field of information science. Research is conducted in close collaboration with commercial partners, particularly in the industrial sector.

The laboratory’s scientific activity is organized around three teams with complementary skills:

o   The CID team (Knowledge, Uncertainties, and Data)

o   The SCOP team (Security, Communication, and Optimization)

o   The SyRI team (Interacting Robotic Systems)

The platforms and demonstrators developed at the laboratory testify to Heudiasyc’s commitment to applying its research to the complexities of real-world applications.

The laboratory has four platforms used in French national innovation programs (Equipex+), supported by research staff:

• Intelligent and autonomous vehicles,

• Aerial mini-drones,

• Rail traffic supervision,

• Virtual reality.

https://www.hds.utc.fr/en/research/technological-platforms/

Contract type and excepted start date

Fixed-term contract – expected duration of 2 to 3 years, with the possibility of extension – to be filled in early 2027

Gross monthly salary

From €2,590 to €5,000, depending on experience and funding

Application dates

From 29/07/2026 to 30/09/2026

Contacts

Yves Grandvalet, CNRS Director of Research, PI of the CAP “AI for Industry and Risk Management” of PostGenAI@Paris

Yves.Grandvalet@utc.fr

Profile and keywords

Profile: Research experience in one or more scientific areas related to the "AI for industry and risk management".

Keywords: Artificial intelligence, AI for industry, computer vision, machine learning, trustworthy AI, explainability, uncertainty quantification, robustness, auditability, AI certification.

Qualification

Required degree: PhD

A PhD degree is not required at the time of application, but it will be required following the interview

Field: Computer science and related disciplines.

The recruited candidate will be expected to:

·        Work collaboratively and contribute to the scientific coordination and animation of the CAP initiative,

·        Show interest in technological research and industrial partnerships,

·        Demonstrate proficiency in English,

·        Disseminate research results through publications and other forms of valorization,

·        Teach diverse student audiences.