Good Practice - Project
3D Photogrammetry and Energy‑Efficiency Innovations in Vertical Farming
Latvian EIP Operational Group advances vertical farming by developing and testing innovative energy solutions helping to enhance yields.
General information
RDP Priority
- P1. Knowledge transfer and innovation
RDP Focus Area
- 1A: Innovation & cooperation
RDP Measure
- M16: Cooperation
Beneficiary type
- Operational group
Summary
This EIP Operational Group (OG) project addressed key challenges in vertical farming in Latvia, particularly the high energy consumption of controlled‑environment systems and the need to optimise environmental conditions to improve crop yields. The initiative was implemented at the Mežvidi greenhouse complex of SIA Latgales dārzeņu loģistika, and brought together an interdisciplinary consortium of experts, researchers, and students. Their combined expertise in agriculture, artificial intelligence, and engineering enabled the development and testing of innovative solutions for energy efficiency and crop monitoring.
Project activities included the introduction of AI‑based plant monitoring using 3D photogrammetry, optimisation of LED lighting parameters, upgrades to the electrical infrastructure, and improvements to climate control in modular vertical farming systems. Wide‑angle cameras were installed to capture real‑time visual data, which was processed by machine‑learning algorithms capable of detecting early signs of plant stress and issuing alerts to agronomists. Custom‑designed lighting systems and tailored ‘light recipes’ were also tested, resulting in improved microgreens performance and demonstrating the importance of uniform lighting and stable climate conditions for consistent crop quality.
Results
- The project successfully demonstrated practical, interdisciplinary approaches to reducing electricity consumption by up to 30%, optimising crop yields, and generating new knowledge for the development of vertical farming in Latvia.
- The work also highlighted the importance of stable climate control and reliable IoT communication, revealing the limitations of wireless sensors in metal containers and identifying opportunities to improve thermal efficiency through alternative climate‑system configurations.
- AI‑based image analysis was successfully piloted in tomato production, showing potential to support agronomists through early‑warning alerts.
- Overall, the project advanced technological understanding in vertical farming, stimulated innovation across the sector, and demonstrated strong interdisciplinary collaboration, including the involvement of emerging specialists and students.
Promoter
Mežvidi Agricultural and Technology Park*
* The Project promoter/beneficiary is an EIP-AGRI Operational Group ©
Funding
Total budget: 343 692.90 (EUR)
RDP support: 311 182.57 (EUR)
Private/own: 32 510.33 (EUR)
Resources
Links
Context
Thanks to rapid technological development, Europe’s vertical farming sector has experienced significant growth over the past decade. Vertical farming is a method of growing crops in vertically stacked layers within controlled indoor environments. It relies on artificial lighting, climate regulation, and efficient irrigation systems to produce food with minimal land use. This approach enables year‑round production, reduces water consumption, and can significantly increase yields, compared to traditional open‑field farming. The global vertical farming market is projected to grow at a Compound Annual Growth Rate of 24.1% between 2023 and 2031.
Despite this strong growth, the sector remains in an early stage of development, and further research is needed to create commercially viable and sustainable vertical farming systems. Key challenges include developing optimal lighting strategies for different crop groups, reducing the high energy consumption of controlled‑environment systems, improving interoperability between engineering components, and overcoming barriers to data collection.
Energy efficiency is one of the most pressing issues, as artificial lighting and climate control account for the majority of energy use in vertical farms. Testing and validating technological innovations in real cultivation environments requires time and long‑term monitoring to generate high‑quality data on their effectiveness. Addressing energy losses can lead to substantial improvements in system performance, while identifying the most suitable lighting parameters for specific crops can significantly increase yields.
Objectives
The main objective of the project was to advance the development of vertical farming by implementing a nationally significant, innovation‑oriented initiative, and introducing new technical solutions to the sector. The specific aim of the project and the vision of the EIP Operational Group was to establish an integrated, highly competent cross‑sector consortium capable of addressing the fundamental technological challenges of vertical agriculture. Its focus was on improving crop yields and reducing electricity consumption through targeted research, technological innovation, and collaborative expertise.
Activities
The OG carried out plant monitoring using the 3D photogrammetry method, which served as the basis for developing innovative methods and technologies to increase yield and energy efficiency in vertical farming.
During implementation, AI-based plant monitoring methods were combined with innovative technological improvements in a modular vertical farm. Overall, the OG combined knowledge and expertise to find solutions to challenges related to artificial lighting, energy efficiency, improvement of climate stability, and data collection through the lens of artificial intelligence.
In addition, a monitoring system was developed, consisting of visual data acquisition of growth parameters and a machine learning algorithm capable of identifying plant condition within the vertical farming system.
Main results
- In container‑based vertical farming, electricity consumption is one of the largest cost components, driven mainly by LED lighting and the cost of the fixtures themselves. Although the market offers many lighting systems, manufacturers cannot reliably predict how plants will perform under specific light spectra. Within this project, cultivation‑specific data were collected under different lighting conditions, including relatively low‑cost ‘white light’ spectrum solutions. New approaches and customised ‘light recipes’ were also tested, and custom‑designed fixtures demonstrated significantly improved results for microgreens cultivation.
- Different plant species and varieties responded differently to the tested light conditions, resulting in variable outcomes. However, a clear trend emerged: excessive illumination (in terms of micromole levels) is unnecessary, and lighting uniformity is crucial for achieving consistent crop appearance, which is an important factor for marketability.
- Another key insight concerns the importance of maintaining a stable climate, which also depends on reliable wireless communication. IoT sensors installed inside metal containers often experience signal interference, creating challenges for real‑time monitoring and control in production environments. From a thermal‑efficiency perspective, climate systems could be improved by integrating heat pumps or hydrocapillary solutions (water‑distribution systems that use capillary action to deliver moisture efficiently to plants), or at least by using two heating, ventilation, and air conditioning systems units, as operating a single unit at its most efficient capacity is rare.
- In 2025, wide‑angle web cameras were installed in the tomato cultivation area of SIA Latgales dārzeņu loģistika at the Mežvidi greenhouse complex to capture periodic real‑time images. These images are processed by a specialised system using artificial intelligence. In practice, the AI is being trained to support agronomists by providing early‑warning notifications when plant health issues or stress conditions are detected.
Key lessons
- The analysis of energy consumption and the search for energy‑efficiency solutions also attracted the interest of other technology manufacturers. As a result, the sector in Latvia generated new ideas for product development, supported by theoretical estimates suggesting that total heat and electricity consumption in vertical farming could be reduced by 30% or more. What remains uncertain is how plants will respond to these new conditions, but this uncertainty will undoubtedly stimulate further research and generate valuable new knowledge.
- Public interest in vertical farming was high, yet even with safety precautions, occasional and unintentional introductions of insects or diseases occurred. This demonstrated that visits to production sites must be managed carefully, as contamination can force an experiment to be restarted from the beginning.
- Despite unforeseen challenges, such as procurement delays and issues with technological interoperability, the project achieved far more within the available timeframe and resources than originally planned. One clear lesson is the importance of allocating periodic reserve time, which allows teams to redirect efforts to other sub‑activities and avoid bottlenecks elsewhere.
- Project implementation coincided with the COVID‑19 period, when communication and in‑person collaboration were severely restricted. The opportunity to observe practical examples on-site and discuss them directly would have provided additional value. Although experts from Luxembourg programmed the AI component, broader exposure to real‑world applications and hands‑on experience with the technologies would have been highly beneficial.
- Overall, the more similar projects are reviewed and compared, the greater the momentum and added value for the sector. Students at various levels of study were involved in the project, and it is clear that strong specialists are emerging in Latvia who are motivated to tackle practical, complex, and interdisciplinary challenges, having also demonstrated an impressive ability to collaborate across different age groups and educational backgrounds.