MSCA PF 2024 EoI-Integrated Building and Rooftop Greenhouse (RTG) Optimization with Machine Learning-based Predictive Control

The current Expression of Interest is intended for searching post-doctoral candidates with interest in applying for a MSCA-2024 Postdoctoral Fellowships call (MSCA Postdoctoral Fellowships 2024 (HORIZON-MSCA-2024-PF-01) at the Universitat Politècnica de Catalunya (UPC, Barcelona) as Host Institution. 

Potential candidates must comply with the following eligibility criteria:

• to be in possession of the PhD title
• to have a maximum of 8 years of postdoctoral experience in research (exceptions to this rule can be found in the call description).  
• Mobility rule: do not have resided or carried out the main activity (work or studies) in the country of the beneficiary (for the modality European Fellowships) or in the country of the Associated party participating as Host Organization during the outgoing phase (for the modality Global Fellowships) for more than 12 months during the last 36 months immediately prior to the call deadline. 

DESCRIPTION OF THE HOST ORGANIZATION AND RESEARCH GROUP

The UPC is a public institution of research and higher education in the fields of engineering, architecture, sciences and technology, and one of the leading technical universities in Europe according to international rankings (Shanghai, QS). Since 2013 the UPC has endorsed the European Commission’s European Charter and Code for Researchers. In the context of the Human Resources Strategy for Researchers (HRS4R), which supports the Charter, the UPC obtained (in July 2017) an official acknowledgment for HR Excellence in Research and was granted the certified logo of the programme by the European Commission.

The Group of Construction Research and Innovation (GRIC) at Universitat Politècnica de Catalunya (UPC) is pioneering research on integrated building and rooftop greenhouse

(RTG) systems. This research delves into the synergies between these systems and leverages machine learning (ML) for advanced modelling and control strategies.

GRIC's approach centres on developing accurate grey box models for both the building and RTG. These grey box models combine physical understanding with data-driven techniques to capture the intricate interplay between environmental variables (temperature, humidity, CO2) and control actions (ventilation, heating/cooling, lighting). This data-driven approach forms the foundation for a powerful Model Predictive Control (MPC) strategy. The MPC strategy will utilize the developed models to optimize building performance in real-time. This translates to achieving multiple objectives simultaneously: maximizing occupant comfort, optimizing RTG crop production, and enhancing energy efficiency. The MPC will be designed for multi-objective optimization, prioritizing factors such as minimizing building energy demand and cost while maximizing RTG crop yield. Machine learning, particularly reinforcement learning, will be a key player in this project. It will enable the system to automatically learn and adapt optimal control policies based on real-time data streams. This data will encompass weather patterns, building occupancy patterns, and crucial crop growth stages.

The project envisions significant reductions in energy consumption and cost for the integrated building and RTG system. This will be achieved alongside improved occupant

comfort and increased crop yield within the RTG. By harnessing the power of machine learning and grey box models, GRIC aims to establish a new paradigm for sustainable

and resource-efficient building-greenhouse integration.

How to apply?

Interested candidates should send their CV and a motivation letter through the EURAXESS portal/contact e-mail provided.

Contact: Prof. Marcel Macarulla Martí