Autonomous Greenhouses – Smart Micro Farming and Large-scale Production

Combining AI and IoT technologies to establish an automated greenhouse management system able to monitor climatic conditions and carry out robotic crop treatments.


This FIE is aware of the large-scale potential of plant cultivation in gardens and greenhouses of Smart Cities and applies Artificial Intelligence (AI) and Internet of Things (IoT) solutions to achieve this objective. Therefore, they will develop a specialised micro-farming solution for individual users which includes several sensors and actuators to fully automate the management of greenhouses. It further requires the identification of key data needed for the efficient use of the system, relying on AI to provide decision assistance. Incorporated user feedback and the implementation of IoT components will additionally help to protect against the influence of external and environmental factors.

As a result, greenhouses will monitor their climatic conditions and carry out robotic treatments accordingly, including soil preparation, sowing, weeding and crop harvesting. Such an intelligent production support system will be able to predict yields at specific settings and help to indicate or eliminate production errors to optimise production costs for each micro-farm.


To reach its goals of end-user adoption the project will leverage the best practices of EU producers of greenhouse sensor systems, measuring bases, data processing as well as business model support and resource planning systems in agriculture. The target audience includes producers of pepper, tomatoes, soft fruits or flowers alongside Smart City enthusiasts and gardening associations.

Lessons learnt:

During the implementation of the pilots, problems with the proper drive transmission were noted when the robot frame was dirty. The susceptibility of engines to increased resistance has been noted, however, it is difficult to eliminate this problem apart from periodic preventive cleaning of the track. The problem only arises if the user intervenes in test field. The problem is solved by raising the awareness of the user. In the end, it was not necessary to change the power of the engines.

A pandemic could unexpectedly and severely affect the component supply chain and component availability. There are also some components that are manufactured exclusively in China and it is impossible and difficult to replace them when the supply chains are interrupted (especially electronic components). In the period of lack of components, their price increases dramatically. Due to the increase in prices, we had to reduce the number of piloted constructions from planned 6 to 4.