Catalogue of Services

This service includes all essential actions to gather knowledge, design, and document the client’s needs regarding Testing and Experimentation Facility (TEF) services. Through close collaboration with the client, we ensure that all technical, operational, and methodological aspects are clearly outlined, forming a solid foundation for future testing phases. By conducting these preparatory activities, we deliver a comprehensive setup analysis that serves as the entry point for subsequent testing and validation services and provision of datasets. This ensures that testing can be carried out efficiently, aligning with the client’s specific requirements while leveraging TEF’s expertise and infrastructure. Ultimately, this service streamlines the testing process, reduces uncertainties, and enhances the overall efficiency of AI and robotic solution validation.

Politecnico di Milano (POLIMI)

Collection of test data during physical testing

Università degli Studi di Milano (UMIL)

Through this service, we offer dedicated technical systems (hardware and software) and expertise for the collection of relevant data during testing.

The scope of data collected can range from data generated by the AgrifoodTEF technical infrastructure (e.g., through the installation of IoT sensors) to the collection of data generated by the system under test to the generation of ground truth for evaluating the system. If necessary, service S00111 may be leveraged to take care of the interconnection between AgrifoodTEF data collection systems and the customer’s systems.

In addition to the raw collected data, we will also provide customers with complete documentation describing the logged features and conditions of the testing environment at the time of testing, as well as any parameter values, variation ranges and specifics required for reproducibility purposes.

Where necessary, data collection and documentation activities can require the involvement of field experts (e.g., engineers, agronomists).

Note: annotation and labelling of collected data, if needed, can be provided via service S00115.

Computer vision for automated food quality assessments

Validation of an existing automated food quality assessment solution based on Artificial Intelligence, in particular Computer Vision techniques. The service will help the solution provider check its performance using a controlled testing infrastructure, including our Food Processing Pilot Plant. The service might also include the generation of reference datasets of food samples and expected quality measures, which can be used both for validation and training of the Artificial Intelligence solution based on Computer Vision.

AI model training

Politecnico di Milano (POLIMI)

Data validation and processing

Università degli Studi di Milano (UMIL)

Through this service, we offerers the ability to validate and augment the data collected during the different testing phases. Examples of operations that this service can provide include quality checks (e.g., presence of all datastreams, presence of drop-outs, technical quality such as focus or exposure of camera images, agronomic significance of data); preparation for performance evaluation (e.g., identification of the images containing selected visual markers or specific plants, association between images from the system under test and the ground truth); data processing (e.g., transcoding); and augmentation (e.g., generation and incorporation of metadata). Data labelling (e.g., to enable training of AI models) is not included as it is the object of a separate service (S00291).

Politecnico di Milano (POLIMI)

Design of evaluation metrics for physical testing

Università degli Studi di Milano (UMIL)

Any physical test activity involves three main components: environment (where the tests take place), protocol (defining what tests are executed and how) and evaluation metrics (used to assess the results of the tests). This service concerns the last element; its goal is to design the best metrics to evaluate the performance of a customer solution, taking into consideration the use cases specified by the customer and the environment and protocol chosen for the tests (which, if needed, can be designed via services S00106 and S00107).

Our team will identify and define with customers the most adequate set of quantitative (i.e., based on instrumental measurements) and/or qualitative (i.e., relying on expert human judgement) metrics to assess the system functionalities of interest. This phase will involve, in particular, agronomists and experts in agricultural machinery.

Based on the defined evaluation metrics, a set of requirements for the collection of required data and ground truth annotations will also be defined accordingly. For instance, the service may lay out the specifications for dedicated data collection campaigns (possibly executed via service S00113). This phase will involve engineers and experts in AI and robotics.

On request, the output of the service will include analyses on additional environmental factors than those directly tracked through the designed metrics (e.g., seasonal effects, impact of test distribution over time on results).

Evaluation of ethical aspects in agrifood technology design and development

We offer a specialised service to companies, particularly small and medium-sized enterprises (SMEs), that develop AI- and robotics-based products and services for the agri-food sector. Our service supports these companies in integrating ethical considerations throughout the design and development process of their AI- or robotics-based products and services. By utilising Value Sensitive Design (VSD) methods, we ensure that human values such as sustainability, fairness, transparency, and privacy are incorporated into technological solutions from the outset. Through stakeholder analysis, participatory design, value scenarios, and ethical reflection, we guide companies in creating responsible, innovative technologies that align with societal needs and ethical standards.

ELSA assessment

National Institute for Research in Digital Science and Technology  (INRIA)

Provision of general-purpose datasets via a multisensory aerial robot.

We provide general-purpose datasets that can be used by customers to evaluate mobility algorithms and to develop and assess general-purpose AI applications. In the context of mobility algorithms, this pertains to classical robotics tasks such as mapping, localisation, and SLAM (Simultaneous Localisation and Mapping). Meanwhile, general-purpose AI applications focus on advancing algorithms and feeding decision support systems (DSS) for tasks including, but not limited to, weed detection, health monitoring, growth and maturity assessment, and yield estimation in areas such as arable farming, horticulture, food processing, forestry, and tree management. A significant challenge in developing AI solutions for agricultural robotics lies in the dynamic nature of agricultural environments, which fluctuate with different seasons and weather conditions. To address this, acquiring consistent and periodic data is essential for effectively monitoring these changes. This real-time data collection, often facilitated by aerial robots, is crucial for developing efficient algorithms and AI solutions. Such datasets can support customers in the development of sensor-specific techniques or be leveraged to create multisensory algorithms, enabling more accurate and adaptable systems for agricultural applications.

National Institute for Research in Digital Science and Technology  (INRIA)

Provision of general-purpose datasets with user-specified sensor(s)

General-purpose datasets serve two primary objectives: (i) evaluating mobility algorithms and (ii) developing and assessing general-purpose AI applications. In the context of mobility algorithms, this includes classical robotics tasks such as mapping, localisation, SLAM (Simultaneous Localisation and Mapping), and navigation. Meanwhile, general-purpose AI applications focus on advancing algorithms and supporting decision support systems (DSS) for tasks such as, but not limited to, weed detection, health monitoring, growth and maturity assessment, and yield estimation in areas like arable farming, horticulture, food processing, forestry, and tree management. A significant challenge in developing AI solutions for agricultural robotics lies in the dynamic nature of agricultural environments, which fluctuate with different seasons and weather conditions. To address this, acquiring consistent and periodic data is essential for effectively monitoring these changes. This real-time data collection, often facilitated by aerial and/or ground robots equipped with user-specified sensors, is crucial for developing efficient algorithms and AI solutions. Such datasets can support the development of sensor-specific techniques or be leveraged to create multisensory algorithms, enabling more accurate and adaptable systems for agricultural applications.

National Institute for Research in Digital Science and Technology  (INRIA)

Provision of general-purpose datasets via multisensory ground robot

General-purpose datasets serve two primary objectives: (i) evaluating mobility algorithms and (ii) developing and assessing general-purpose AI applications. In the context of mobility algorithms, this pertains to classical robotics tasks such as mapping, localisation, SLAM (Simultaneous Localisation and Mapping), and navigation. Meanwhile, general-purpose AI applications focus on advancing algorithms and feeding decision support systems (DSS) for tasks such as, but not limited to, weed detection, health monitoring, growth and maturity assessment, and yield estimation in areas like arable farming, horticulture, food processing, forestry, and tree management. A significant challenge in developing AI solutions for agricultural robotics lies in the dynamic nature of agricultural environments, which fluctuate with different seasons and weather conditions. To address this, acquiring consistent and periodic data is essential for monitoring these changes effectively. This real-time data collection, often facilitated by ground robots, is crucial for developing efficient algorithms and AI solutions. Such datasets can support the development of sensor-specific techniques or be leveraged to create multisensory algorithms, enabling more accurate and adaptable systems for agricultural applications.

National Institute for Research in Digital Science and Technology  (INRIA)

Testing and evaluating sensor devices

The SOPHIA infrastructure will offer the possibility to test and evaluate the performance of a sensor device in a locally controlled environment. While our indoor facility provides a fully controlled environment (light, volume), all facilities in SOPHIA can be customised to test and evaluate the sensor device. The service consists of three main steps. First, the sensor device will be integrated into the facility. After that, the facility will be customised according to the final application of the sensor device. Then, the performance of the sensor device will be evaluated using quantitative and qualitative metrics. Comparison could be proposed as a complementary option to position the performance of the sensor device in relation to the current state of the technology available in SOPHIA.

People training

National Institute for Research in Digital Science and Technology  (INRIA)

Testing and evaluating the robot’s functionalities

The SOPHIA infrastructure will offer the possibility to test and evaluate the performance of the robot’s functionalities in a locally controlled environment. While our indoor facility will offer a fully controlled environment (light, volume), all facilities at SOPHIA can be customised to test and evaluate the robot’s functionalities. The service consists of three main steps. The robot will be integrated into the facility. After that, the facility will be customised according to the functionality of the robot. Then, the performance of the robot’s functionalities will be tested and evaluated with quantitative and qualitative metrics.

AI model training

People training

National Institute for Research in Digital Science and Technology  (INRIA)

Testing and evaluation of mobility algorithms with ground robots

The SOPHIA infrastructure provides the ability to test and evaluate mobility algorithms embedded on a ground robot. Mobility algorithms concern the classical robotics functionalities of mapping, localisation, SLAM, and navigation. The ground robot is equipped with an array of sensors, including a camera, LiDAR, IMU, and RTK-GPS for ground truth evaluation. The service proceeds in three stages. Firstly, we evaluate the algorithm using representative datasets. After that, the algorithm is integrated into a ROS2 architecture and evaluated with the local agrifoodTEF test infrastructure (various areas are possible). The performance of different attributes of the algorithm is assessed using quantitative and qualitative metrics. Benchmarking could be proposed as a complementary option to position the performance of the proposed algorithm in relation to the current state of the art. The final step involves field testing under real conditions at a specific end-user or customer site using the mobile living lab, which consists of a mobile laboratory deployed in the field and connected to the real robot for monitoring and evaluation purposes.