The recent emergence of plant factories highlighted the need for automated plant cultivation in fully controlled environments. The automation enables the determination of ideal plant varieties and cultivation parameters to ensure optimal production yields. GDTech, an engineering service company operating out of Belgium and France, along with ULiège, a Belgian university, recognized the potential of these plant factories as tools for their research and development pursuits.

In the process of designing and implementing these advanced automated plant factories, our team worked closely with key stakeholders in the project, integrating the hardware and software components of the solution. The hardware infrastructure involves robotic equipment along with IoT sensors, actuators, and cameras, which collect telemetry and imaging data from the cultivation unit. The setup enables us to create an efficient cloud-based infrastructure, leveraging AWS IoT related services, that is tailor-made to support the hardware. The end goal here is to explore the feasibility and efficacy of automated cultivation for specific plant varieties in this fully regulated environment. The infrastructure we developed harnesses AWS IoT services to collect data from robotic actuators, environmental variables, and hyperspectral camera imagery, condensing it into a scalable framework. The collected data is subsequently analyzed to digitally phenotype the plants under cultivation and to autonomously determine the optimal cultivation parameters. Fully remotely controlled by researchers, this plant factory unit offers the ability to monitor collected data via interactive dashboards, allowing for a comprehensive view of the automated cultivation processes in real time.

We conceptualized and executed a comprehensive cloud-based platform for the effective monitoring, management, and control of a plant factory unit, with the ultimate aim of enhancing plant production. The solution is currently used for research and development purposes and aims at optimizing industrial processes in future developments.