Intelligent IoT System for Greenhouse Automation (SIOTIN), funded by UEFISCDI under the PTE Transfer to Economic Operator axis for the period 2025 – 2027.
SIOTIN aims to develop a greenhouse automation product to support sustainable agriculture. The project’s objective is to optimize resource usage and enhance both productivity and crop quality in greenhouses through state-of-the-art technologies. It will develop an integrated monitoring and control system consisting of:
- Multisensor nodes for monitoring microclimate parameters,
- Actuator nodes for controlling greenhouse processes,
- A base station for communication between nodes,
- A web platform integrated with a mobile application, operating autonomously through an intelligent decision-making process.
Contract No. 35PTE/2025
- Program Name (PV IV): Program 5.7 Partnership for Innovation
- Subprogram Name: Subprogram 5.7.1 Partnerships for Competitiveness
- Project Type: Transfer Project to Economic Operator
- Project Title: Intelligent IoT System for Greenhouse Automation
- Total Contract Value: 1,810,000 RON
- Source 1 (State Budget): 1,500,000 RON
- Source 2 (Other Sources – Co-financing): 310,000 RON
- Contract Duration: 24 months
- Coordinator: BEIA CONSULT INTERNATIONAL
- Partner 1: National University of Science and Technology POLITEHNICA Bucharest (UNSTPB – upb.ro), CAMPUS Research Institute
- Partner 2: National Institute for Research and Development of Machinery and Installations for Agriculture and the Food Industry (INMA Bucharest)
Project Duration and Activities
The project has a duration of two years (24 months). According to the competition timeline, the project will begin in September 2024 and will consist of three main activities:
Activity 1: Requirements Definition and Technology Selection
Phase I – Requirements definition, technology selection, development of the sensor node, actuator node and base station, and integration of automation algorithms
A1.1 – Requirements definition
Activity A1.1 enabled the complete definition of the functional and non-functional requirements of the SIOTIN system, starting from the real needs of farmers and the limitations identified in current greenhouses. The critical parameters to be monitored were established (air/soil temperature, air/soil humidity, CO₂, light), as well as the specific requirements for all subsystems: sensor nodes, actuator nodes, base station, and intelligent algorithms. This set of requirements forms the technical foundation for the subsequent activities, ensuring architectural coherence and alignment of the solution with the needs of the end user.
A1.2 – Technology selection
Activity A1.2 resulted in the selection of the most suitable hardware and software technologies for the implementation of the SIOTIN system. Robust solutions were identified for sensors, acquisition boards, communication protocols, and radio modules, with selection criteria including accuracy, energy consumption, compatibility, scalability, and resistance to agricultural environments. The outcome is a modular, flexible, and extensible hardware architecture capable of supporting the autonomous and safe operation of the entire IoT system.
A1.3 – Hardware and software development of the sensor node
Activity A1.3 led to the development of a fully functional sensor node optimized for acquiring the critical parameters of the greenhouse microclimate. The use of the Arduino MKR WAN 1310 platform together with a customized sensor shield enabled the creation of a low-power device capable of autonomous operation and reliable data transmission via LoRaWAN. The implementation of selective powering for each sensor improved energy efficiency, while local processing (filtering, validation, handling of missing values) ensured the quality of the transmitted data. The resulting sensor node meets the robustness, accuracy, and modularity requirements necessary for the implementation of the SIOTIN system.
A1.4 – Hardware and software development of the actuator node
Activity A1.4 resulted in the development of a versatile and reliable actuator node capable of controlling essential greenhouse equipment such as irrigation, ventilation, lighting, or shading. The architecture based on the Arduino MKR WAN 1310, combined with an electrically isolated and protected relay module, ensures operational safety and compatibility with multiple types of loads. The node supports both local and remote commands, LED-based diagnostics, and expansion to up to 8 channels. Through this design, the actuator node offers a high level of flexibility, reliability, and seamless integration into the SIOTIN IoT ecosystem.
A1.5 – Development of the base station
Activity A1.5 enabled the creation of a robust base station capable of managing bidirectional communication with the sensor and actuator nodes, as well as running intelligent control algorithms. The Raspberry Pi 4 platform combined with the RAK2245 LoRaWAN concentrator module ensures excellent radio coverage and adequate processing capabilities for local inference. The base station was integrated into an IP67 enclosure, ensuring protection and stable operation in agricultural environments. It is designed to operate autonomously even in the absence of internet connectivity, which is a critical requirement for a reliable agricultural IoT system.
A1.6 – Development of intelligent algorithms
Activity A1.6 established the foundations of the SIOTIN system’s decision-making mechanism, combining threshold-based agronomic control with a Random Forest model for intelligent automation. Procedures for data filtering, normalization, and gap-filling were defined, and a synthetic dataset of 5,000 records enabled the initial testing and calibration of the model. The RF algorithm demonstrated the ability to handle complex microclimate variations, offering adaptability, noise tolerance, and the capability to predict critical situations. This activity lays the groundwork for a proactive and autonomous greenhouse control solution.
A1.7 – Algorithm integration
The integration of the algorithms into the complete SIOTIN system architecture transformed the platform into a functional system capable of monitoring, analyzing, and acting autonomously. Implementing the RF logic directly on the base station ensures low reaction times and independence from internet connectivity. The established operational process—acquisition, preprocessing, inference, encoding, and transmission—enables continuous microclimate control and resource optimization. This activity confirms the maturity of the architecture and the compatibility between all components.
Conclusions of Stage I
Stage I of the SIOTIN project established the complete technical foundation for the development of an intelligent IoT-based greenhouse automation system. The requirements were defined, optimal technologies were selected, the sensor and actuator nodes were developed, the base station was implemented, and the control algorithm was integrated. In parallel, scientific results were published in international conferences. The stage concludes with a coherent, conceptually tested system, ready for field implementation and for the experimental activities planned in subsequent stages
Dissemination
D. Săcăleanu – In-field environmental parameters monitoring, AI4AGRI Summer School 2025 – EO Big Data for Agriculture, 14 – 19 July 2025, Brasov, Romania
D. Săcăleanu et. all, PRECISION AGRICULTURE EXTENSION MODEL BASED ON INTERDISCIPLINARY COLLABORATION, 14th Edition of the International Conference “Agriculture for Life, Life for Agriculture”, Bucharest, 5-7 June 2025
- Matache, M.-G., Cristea, R., Sîcleanu, D.-I., & Dobre, C.-M. (2023). Design and simulation of a random forest-based control algorithm for greenhouse management. International Symposium, 232–240. INMA Bucharest; National University for Science and Technology Politehnica Bucharest; Beia Consult International.
Deliverable:
- D1. Technical Note on Requirements and Technology Selection (T4)
This document will detail the overall use-case scenario, the system’s technical requirements based on farmers’ needs, and the hardware/software technologies selected for system development.
Activity 2: Development of the Integrated System
Objectives:
- Develop sensor nodes, actuator nodes, and the base station.
- Develop and integrate intelligent automation algorithms.
- Develop the web platform and mobile application.
Sub-activities:
- A2.1 Hardware & Software Development of Sensor Node (T4 + 6 months) – Politehnica University of Bucharest
Wireless sensor nodes will be developed, including schematic design, PCB creation, assembly, and sensor integration. An alternative energy source using photovoltaic panels will be implemented. The sensor node will be enclosed in a sealed case suitable for greenhouse use, with general output connectors. It will support flexibility for attaching various sensors based on farmers’ needs. The software will handle data acquisition, processing, and transmission, with energy-saving algorithms implemented. - A2.2 Hardware & Software Development of Actuator Node (T4 + 6 months) – Politehnica University of Bucharest
Actuator nodes will be developed similarly to sensor nodes. The casing will include cable glands for wiring. Power supplies will be adapted based on the required equipment (24VAC, 230VAC). The software will process incoming commands and control equipment accordingly. Commands can be manual or automatic, with local execution plans received from the base station. - A2.3 Hardware & Software Development of Base Station (T8 + 8 months) – Politehnica University of Bucharest
The base station will be built using a LoRa concentrator and a development board with internet connectivity. Its software will manage data reception from sensors, command transmission to actuators, and two-way communication with the web platform. - A2.4 Development of Intelligent Automation Algorithms (T9 + 5 months) – Politehnica University of Bucharest
Automation algorithms will be based on threshold values configured by the farmer via the web platform and will also include learning and prediction capabilities for proactive control. Emergency and alert protocols (e.g., automatic window closure during strong winds) will also be defined. - A2.5 Integration of Automation Algorithms into the System (T13 + 3 months) – BEIA
Intelligent algorithms will be integrated into the base station to enable independent operation after the personalized crop plan is configured. - A2.6 Web Platform Development (T8 + 8 months) – BEIA
A platform will be developed for monitoring and controlling greenhouse microclimatic parameters. It will provide personalized dashboards per farmer, allowing configuration and real-time visualization of parameters, system automation settings, and manual control options. The platform will be hosted on BEIA’s server. - A2.7 Mobile Application Development (T10 + 6 months) – BEIA
A mobile app will allow farmers to access their equipment database, view parameter trends and real-time values, and control actuators. The app will offer both manual and automatic control modes.
Deliverables:
- D2. Technical Note on Wireless Sensor Network Development (T16)
Detailed description of hardware and software components, including schematics, PCBs, enclosures, and software codes (as appendices). - D3. Technical Note on Automation Algorithm Development & Integration (T16)
Complete description of the developed algorithms, including implementation codes. - D4. Technical Note on Platform Development (T16)
Detailed overview of the web and mobile platforms, development methodology, and functionality.
Activity 3: System Integration, Testing, and Validation
Objectives:
- Integrate, test, and validate the system.
- Disseminate project results.
Sub-activities:
- A3.1 System Integration (T16 + 2 months) – BEIA
All hardware and software components will be integrated into the final system. - A3.2 Laboratory Testing & Validation (T18 + 2 months) – BEIA
Functional testing of connectivity and automation features will be performed in the lab. Sensor readings and actuator actions will be validated using calibrated and certified measurement equipment. - A3.3 Greenhouse Testing & Validation (T20 + 4 months) – BEIA
The complete system will be tested in real conditions in a greenhouse owned by partner INMA. Various sensor and actuator configurations (e.g., electrovalves, fans, motors, lighting, water pumps) will be validated. - A3.4 Project Dissemination (T16 + 8 months) – BEIA
This activity aims to inform public and private stakeholders about the research topics and project outcomes.
Deliverable:
- D5. Technical Note on System Integration, Testing, and Validation (T24)
This report will detail the integration process, lab and real-environment tests, and the results obtained for validation.
The Kick Off Meeting took place on 18th of March 2025.
