Having precise and immediate information on plant health is crucial for planning targeted and effective interventions aimed at optimizing the distribution of water and plant protection products, thereby reducing costs, safeguarding crops, and increasing productivity.

In situations such as critical weather events, drone-based analysis may also represent the most immediate and effective solution for obtaining an accurate overview and assessing the most appropriate actions to implement.

In the case presented, multispectral data acquired using the MicaSense Altum-PT sensor mounted on the MAVtech RX3 UAV platform were processed to monitor a traditionally trained trellised vineyard.

High-resolution multispectral and thermal images, subsequently processed through a Structure from Motion (SfM) photogrammetric workflow, enabled the generation of a georeferenced orthomosaic, a digital elevation model (DEM), a dense point cloud, and a 3D vineyard model—essential tools for morphological and vegetative analysis of the crop.

Specific vegetation analysis algorithms were then applied to calculate the Leaf Area Index (LAI) and the MCARI/OSAVI indices, useful for assessing the physiological state of vegetation, vegetative vigor, and chlorophyll response.

By comparing the NDVI (Normalized Difference Vegetation Index) and the NDRE (Normalized Difference Red Edge Index), differences in sensitivity in estimating vineyard vegetative status were highlighted. In particular, NDRE is more effective in analyzing crops with high leaf density, proving more sensitive to variations in chlorophyll content during the later stages of the vegetative cycle compared to the traditional NDVI.

In robotics and drone applications, Vicon’s ultra-precise indoor localization system is the preferred solution when control and navigation algorithms must be validated with extreme accuracy, including swarm configurations, as it can simultaneously track the position of multiple drones and robots.

In collaboration with GPEM, a distributor and technology integrator of Vicon systems in Italy for more than 10 years, MAVTech has developed solutions enabling its drones to use positioning data provided by the system either through a direct local network connection (without the need for onboard processing) or through computations performed directly onboard the drone.

The solution, successfully tested and deployed at the “Futura Innovation Lab” of TXT Group, makes it possible to easily operate multiple drones simultaneously in indoor environments, providing an effective interface for developers working on navigation algorithms, autonomous flight, multi-drone systems, and UAV cooperation.

We are also working together on further developments to improve the integration between our drones and the Vicon system, with the goal of making the user experience increasingly simple, intuitive, and effective.

We are proud to have launched this collaboration between MAVTech and GPEM, aimed at providing advanced solutions for universities and public and private research institutions, thereby contributing to technological development in the drone industry.

MAVTech systems, designed and manufactured in Italy, integrate hardware, software, and communication platforms to support efficient operations in inspection, monitoring, surveying, search and rescue activities. The article also highlights new developments currently underway, including VTOL platforms, advanced swarm-based systems and AI algorithms, indoor drones for confined environments without GPS, and collaborative solutions between our drones and the rover systems of partner Info Solution, aimed at continuous monitoring and safety in complex industrial environments such as those in the Oil & Gas sector.

Founded as a spin-off of the Politecnico di Torino, MAVTech continues to develop innovative solutions, transferring research know-how to the market, with applications increasingly focused on safety, operational efficiency, and sustainability.

To learn more: click here.

Access more articles here.

Download the full magazine.

Contact us: mavtech@mavtech.eu

Key points

The event confirmed that the technology developed within START2 is not only technically sound, but also operationally relevant. The significant results achieved by all partners strengthened our confidence in the real-world applicability of the living lab approach — particularly in mountain rescue and emergency response scenarios.

For MAVTech, the most meaningful outcome was the positive feedback on our tethered drone use case, which enabled the restoration of TETRA radio coverage in a signal-deprived canyon during the Pilot Day 2025 at the Serrai di Sottoguda.

Results achieved

MAVTech presented the results of eight development streams carried out within START2: the evolution of the Q4X drone platform (increased endurance, RTK precision, and progress toward EASA C3/C5 certification), the new compact multi-mission RX3 drone (EASA C2 certification pathway), integrated tethered drone solutions for long-endurance operations, wildfire detection systems using drone-mounted sensors, a preliminary vertiport design for rapid emergency response, the VTOL Lifeseeker for long-range search-and-rescue missions, and an automated mission planning tool for hiking trail operations.

The exchange with rescue professionals, institutional partners, and other technology developers confirmed a shared vision regarding the integration of UAVs into emergency operations in alpine environments.

Throughout the project, MAVTech was proud to share updates, results, and insights via digital channels such as LinkedIn, Instagram, and Facebook, reaching both its professional network and a wider audience interested in drone technology and civil protection.

In addition, participation in Civil Protect 2025 and REAS 2025, due tra le più importanti fiere in Italia dedicate alla protezione civile e al soccorso d’emergenza, ha permesso a MAVTech di presentare e promuovere i risultati del progetto START2 tra le comunità di utenti finali più rilevanti.

Equipped with a LiDAR sensor and an RGB camera mounted on a steerable gimbal, the MAVtech UAV Q4X flight platform reconstructed the three-dimensional structure of the infrastructure with very high precision, also demonstrating excellent position-holding performance while hovering, despite wind gusts of up to 5 m/s.

The test was planned in compliance with safety requirements and current regulations and took place in two separate locations in the municipality of Gargazzone.

To learn more: click here.

MAVTech will develop an innovative solution of autonomous drones capable of operating in complex and confined environments, such as chemical plants, tunnels, penstocks, and geological sites, where the GNSS signal is absent or degraded. These contexts represent a challenge for conventional drones due to narrow spaces, unknown obstacles, and variable lighting conditions.

The project aims to create a fully autonomous UAV, capable of localizing itself, mapping the environment, and avoiding obstacles in real time, providing precise and safe data for industrial surveying missions, infrastructure monitoring, and Search & Rescue.

Thanks to its consolidated experience in the design and integration of autonomous systems, MAVTech will develop lightweight sensors, software optimized for onboard processing, and interchangeable payloads, ensuring precision, reliability, and versatility.

MAVTech drones will operate in extremely challenging environments, from underground tunnels to complex industrial plants, and will be used for monitoring missions, structural surveying, and safety, thanks to sophisticated Visual Inertial Odometry (VIO), Lidar Inertial Odometry (LIO), SLAM, and obstacle avoidance algorithms developed in-house.

The project represents a fundamental step to strengthen MAVTech’s position in the autonomous systems sector for monitoring, maintenance, and safety in high-risk environments.

To learn more: click here.

ARGOS (Autonomous UAV for Inspection and Monitoring in Confined Environments)

The project aims to develop an innovative Unmanned Aerial Vehicle (UAV) platform for inspections and surveying in complex industrial and natural environments, such as chemical plants, tunnels, penstocks, and geological sites. These contexts are characterized by confined spaces, lack of GNSS/GPS signals, variable lighting conditions, and obstacles that are not known in advance, making traditional inspection solutions ineffective and limiting the use of conventional drones.
The proposed solution will enable autonomous missions for surveying, infrastructure monitoring, and Search & Rescue operations, even in the absence of external localization infrastructure, ensuring safety and reliability of operations. 

Objectives

The project aims to develop an autonomous UAV capable of operating in confined environments without GNSS signal, with the following main tasks:

1. Autonomous localization: estimation of the UAV’s position using only onboard sensors, with integration through an Extended Kalman Filter. Two complementary approaches will be employed:
   • Lidar Inertial Odometry (LIO)
   • Visual Inertial Odometry (VIO)
     with the goal of achieving high accuracy and reliability beyond commercial standards.
     
     
2. Real-time mapping: creation of incremental 3D maps of the surrounding environment based on advanced SLAM algorithms, optimized for the selected sensors.
3. Obstacle avoidance: dynamic updating of the flight trajectory in the presence of obstacles, ensuring safety and successful completion of the overall mission.

Key Development Aspects:

• Accuracy: reduction of maneuvering margins and access to confined spaces through detailed mapping.
• Reliability: use of redundant and complementary sensors to ensure performance across different scenarios.
• Robustness: testing in environments with variable lighting, reflective surfaces, or non-standard features.
• Software speed: optimized processing for high update rates and reduced energy consumption.
• Versatility: interchangeable payload with different sensors for heterogeneous missions.

The project integrates hardware and software in an incremental approach, from sensor selection and calibration to the development of VIO, LIO, SLAM, and obstacle avoidance algorithms, culminating in experimental testing in controlled environments and real-world scenarios to validate performance against commercial solutions.

Project Details

Title: Autonomous UAV for Inspection and Monitoring in Confined Environments
Acronym: ARGOS
Funding Program: Autonomous Province of Bolzano – Simplified Experimental Development Projects – Act No. 66/2025
Total Budget: €50,000.00
MAVTech Budget: €50,000.00
Duration: January 2026 to December 2027

For this activity, our Q4X drone was prepared and equipped with an ARTVA device (Avalanche Victim Search Device) integrated on an extendable arm. The objective of the test was to verify the system’s functionality under real conditions. The activities carried out provided significant technical insights, allowing us to proceed with the next development steps.

In the upcoming phases, further field tests will be conducted and the system will be progressively optimized, with the aim of making it increasingly effective in supporting search and rescue operations in mountain environments.

Developed in collaboration with Eurac Research, the PARSIVAL project is funded by the Autonomous Province of Bolzano and the European Union under the FESR 2021–2027 program.

PARSIVAL (Aerial Platforms for Avalanche Search and Rescue)

Timeliness in avalanche Search and Rescue operations is of paramount importance for victim survival: the probability of rescuing alive a fully buried avalanche victim is around 80–90% in the first 15–20 minutes after burial, decreasing very rapidly thereafter. After 35 minutes, the survival probability drops drastically to about 30%.
This means that organized rescue teams often fail to arrive in time, and the survival of those victims depends almost exclusively on the expertise of the group members on site and, nonetheless, on the size and slope of the terrain affected by the avalanche.
The solution proposed in the PARSIVAL project will increase the chances of avalanche survival by promoting self-rescue, minimizing the time to locate buried victims, and reducing the probability of human error.

Objectives

The project involves the development of a lightweight (<1 kg), compact, and easily backpack-portable drone that, autonomously guided by a computer vision algorithm based on data from optical sensors and a miniaturized ARTVA device (avalanche beacon), can automatically locate any buried victim in an assisted rescue configuration and report the location to the 112 emergency center or to a search and rescue service.
If drone-guided self-rescue operations fail, further organized rescue actions can be taken, using multiple drones (equipped with optical and thermal sensors) flying simultaneously in swarm configuration.

To achieve this goal, the project activities include the development of:

• A small, lightweight, foldable, and portable aerial platform
• A miniaturized ARTVA sensor carried by the aerial platform
• A computational algorithm to define the boundaries of the search area and autonomously locate the victim(s), even on steeply sloping terrain
• A data support system (DSS) able to communicate between users and rescue operators (SAR). The DSS will coordinate other platforms (ground teams, helicopters, drones) to complete the rescue process

The developed solution will be validated through a field study, so to prove its effectiveness compared to traditional search methods and to evaluate its impact on the performance of companions and rescuers.
The project will contribute to sustainable development as promoted by the European Green Deal: it will improve the efficiency of search and rescue operations through low-environmental-impact platforms (electric-powered) and will reduce the use of more polluting platforms (helicopters) when not required.

Project Details

Title: Aerial Platforms for Avalanche Search and Rescue
Acronym: PARSIVAL
Project Code: EFRE1110
CUP: B57H25003290007
Funding Program: ERDF 2021–2027
Total Budget: €400,786.43 (Funded: €321,835.20)
MAVTech Budget: €197,378.08 (Funded: €118,426.85)
Partner: Eurac Research
Duration: 07/01/2026 – 12/31/2028

The project is structured into several phases, with initial laboratory tests followed by trials in real operational environments. In both scenarios, fleets of drones will be deployed, with one drone dedicated to inspection tasks and multiple support drones providing precise positioning through Ultra Wide Band (UWB) radio technology. The number of support drones will be progressively reduced from an initial 4 to 3 and then 2, in order to evaluate and compare the performance of different configurations and identify the optimal solution.
In addition to the hardware and software development of the drones, MAVTech will contribute to the market analysis for this application, with the aim of defining a viable operational solution for infrastructure inspection.

MAVTech is proud to take part in such a significant project, alongside important partners such as TXT e-solutions (Lead Partner), Fondazione LINKS, Inspection Drone, and the City of Turin, with the support of the European Space Agency (ESA) programme.

For a complete overview of the project, we report here an excerpt from the Dire.it – TG Hi-Tech, available in full at the following link: Friday, December 5, 2025 edition.