top of page
Anchor 1

Research

The unique combination of knowledge and expertise in plant physiology, plant nutrition (i.e., soil and foliar fertilization), climate modelling with ICT and Robotics, is the major strength of this CoLAB leading to multidisciplinary solutions to tackle present and foreseen challenges in farming production systems.

​

The research priorities for the the area of influence of SFCoLAB are related with:

​

  • Better Production & Quality

  • Improved food safety

  • Good control of pests and diseases, and

  • Efficient use of soils.

 

To achieve this under the specific challenges of small size properties and low technological innovation, the development and implementation of new technologies became a key issue. Therefore, the research and innovation agenda will be focused on the smart use of resources to create smart products using smart equipment, and decision tools. Advanced agricultural machinery solutions can help farm holdings – regardless of their size – to operate in a profitable, competitive and sustainable manner.

​

The driving force behind innovation comes from the companies that are part of the CoLAB ecosystem.

Agenda

SFCOLAB Research Agenda

1.

Resources usage efficiency (Water, Energy, Nutrients & Fertilizers) for intensive Small- and Mid-scale Farming

2.

Address current challenges of climate change for mitigation of the impacts of drought and extreme events

3.

Integrated approaches for traceability, certification of origin and product quality

4.

Improving interoperability through standardisation (for example: ISOBUS)

5.

Design and development of Smart Machinery with different levels of autonomy using  different levels of cognition (Artificial Intelligence)

7.

Develop systems and strategies that reduce the use of *icides, fertilizers, water, and carbon footprint boosting ecologically sound farming

8.

Develop energy systems able to utilise the natural energy sources (e.g. by-products from the current process) and provide continuous operation

9.

Developing and applying sensors to get reliable and accurate information aboutspatial and temporal conditions related with soil, weather, and crops.

10.

Developing and applying Artificial Intelligenge (Machine Learning, Big Data and Data Mining) to provide the farmer with access to broader data analysis, including comparative analysis

11.

Design and Development of SMART Applications to help decision making at the different crop's life cycle phase

SFCOLAB Main Research Areas

SMART USE

Toachieve its research agenda and goals the SFCOLAB is organised in 5 research lines.

SMART USE

The SMART FARM CoLAB aims to develop a Sustainable Resource Management framework integrating precise maximization of key resources (including soil) usage efficiency, towards neutral carbon and zero waste farming systems resilient to climate changes (SMART USE):

  1. Increase water, nutrients and fertilizers use efficiency through a SMART real-time framework for improvement of production resilience and sensorial quality (e.g. taste and aroma)

  2. Minimize water losses along all process and within the Fruit- and horti- farming

  3. Optimization of crop irrigation and sustainable drainage water management

  4. Valorisation of co/by resources for a zero waste farming and adoption of the best
    wastewater treatment and recycling water systems in hydroponics use

  5. Implement sustainable, carbon neutral energetic strategies

  6. Adopting bio-fertilizers and -control methods aiming to eliminate chemical traits in crop production.

  7. Keeping alignment with the strategic approach to EU agricultural research and
    innovation agenda such as the European Food Safety Authority Strategy 2020 and RARA agenda, and the National research programme Portugal 2030

​

The main tasks for SMART USE are:

​

Plant Responses - Establishment of the main constraints related to plant responses to environmental stress, in particularly to drought and temperature above and below ground, and optimization of water, nutrients and fertilizers.


Identification and characterization of waste production products - Identification of best wastewater treatment systems and technologies to recycle water in the hydroponics based use; Effluent treatment/mitigation from greenhouse intensive vegetable production as well as re-use of organic waste from greenhouse intensive horticulture (eg, for biodiesel production).


Phytosanitary characterization of resources and co-resource - Aims to disclose biocyber markers for real time diagnostic of the phytopathogenic bacteria and fungi (pests and diseases) and assessment of novel biological processes for sanitization of waste and co-resources.

 

Development of High-throughput plant phenotyping - For real-time trait monitoring biotic and abiotic impact of resources use (including field phenotyping and low-cost phenotyping) for trait monitoring (yield, disease, and other stress factors).

 

Implementation of bio-fertilizers and soil protection measures for sustainablefarming systems -  Using aquaponics, aeroponics, complementary systems - vermicomposting / aquaponics / aquaculture) as well as personalized supplementation of macro and micronutrients from natural sources (algae / Worm compost / other by-products).

 

Sustainable Strategies for mobility in farming systems - Identification of the most suitable energy sources to for cargo transport and farming production system. These analyses can be exploited using a multicriteria analysis and/or multiobjective optimization and is aligned with the strategic axis related with the economy decarbonization.


Energy Audits - To promote the energy consumption characterization, the identification of energy saving opportunities and the implementation of a rational energy usage plan in the farming sector.

SMART EQUIPMENT

The main goal of this WP is doing Research and Innovation activities about SMART Intelligent Farm Machinery to support precision soil preparation, seeding, crop management (fertilization, crop protection, and irrigation), and harvesting. This is the equipment necessary to implement the 4 Rs in agriculture from a squared Km to a Square m level: doing the Right thing, in the Right place, the Right way, at the Right time.

SMART Equipment includes tractors with different levels of autonomy, their intelligent tools (e.g. precision sprayers, combine and specialist harvesters), and Unmanned Aerial Vehicles for remote sensing.

This research line is focused on the design and/or adaptation of the mechanical platforms together with the necessary intelligence that both vehicles and tools must possess. It also focuses on the collaboration between all entities that compose the farming ecosystem. Concretely, the collaboration between SMART machinery, legacy equipment and human workers.

​

The main tasks for SMART EQUIPMENT are:

​

SMART Farming Robots - Unmanned Aerial Vehicles (UAV) and Unmanned Ground Vehicles (Autonomous Tractors) - To adapt and further develop autonomous robotic systems, aerial and land vehicles, to support efficient farming operation, while reducing burden on human actors, including optimization of fertilization and pest combat routes, efficient irrigation, specialized harvest etc. This activity covers all aspects related to perception and autonomy development needed for the SMART Robots. Creating flexible field robots composed of tested building blocks, which allow tailoring robots for all type of applications in farming. Concretely, equipping the robots with adequate decisional autonomy to support continuous farming processes including resource allocation, scheduling, optimization and control. This task is focused on organizing competences around this topic and creating guidelines and mechanisms for design, development and assessment of the following sub-tasks:

Multimodal Perception System - Several sensors and sensor combinations will be tested for autonomous navigation. It is envisioned that the sensor packaged will include a combination of GPS, GPS-RTK, Long and short range multibeam Lidars, Sonars, Stereo Cameras and weather stations.

Advanced Navigation - The goal is to enable the UAVs and UGVs to locate themselves and autonomously navigate during the farming operation. Land vehicles locomotion modules need to permit traction in difficult, heterogeneous, and uneven terrains. The autonomous navigation will rely on a robust data-fused localization, integrating not only on internal feeds but also on external feeds (other vehicles). Also, in environments with limited space like orchards, visual odometry and LiDAR-based localization algorithms can be exploited. Both aerial and land vehicles kinematic motions can then be optimised to reduce both time and energy expenditure and other relevant factors like soil compaction.

Autonomous Mission - While mission planning deals with higher-level tasks that a robot must execute to achieve its goal, motion planning deals with reasoning of the robot’s configuration within the physical world. Furthermore, mission planning should optimise each vehicle operation in terms of resources spent and minimize in an intelligent way their footprint on the field (e.g.: soil compaction, damage on crop covering, etc.). This task is focused on organizing competences around this topic and creating guidelines and mechanismsfor design, development and  assessment of farm related autonomous missions, and training.

​

Development of SMART Cognitive Cyber Physical Tools - The main goal of this task is to develop intelligent farming tools that can be coupled to SMART Robots or legacy equipment. These are Cognitive Cyber Physical tools specifically developed to efficiently execute agricultural related activities, such as a tool developed for smart irrigation or spraying of fertilizers to be attached to a tractor. Tools will exploit both their sensing abilities as well as the data available in the developed framework. Increasing their decisional autonomy to support the farming processes. A good example of these tools is precision sprayers. When used in combination with autonomy, geo-mapping data, the positioning receivers and sensors on the sprayer candetect where weeds and diseases are present and apply the a dequate rates for different areas only as and when necessary. Other examples of Cognitive Cyber Physical tools are crop-specific precision harvesters that exploit mechatronic, perception, and cognitive modules to ensure perfect crop extraction while maintaining the physical integrity of the soil. This task is focused on organizing competences around this topic and creating guidelines and mechanisms for design, development and assessment of SMART Cognitive Cyber Physical Tools, and training.

​

Active Shared Perception for SMART Agriculture Cyber Physical Devices - This task will undertake the appropriate selection of active sensor fusion strategies, required for combining perception data from all entities in the SMART Agriculture Cyber Physical Ecosystem to increase the overall awareness in order to create shared perception mechanisms improving agent understanding of their surroundings, as well as of their own status. The focus will be given on formulating distributed sensor fusion techniques for integration of data from heterogeneous Cyber Physical Devices. Since data can be fused at a variety of levels - from raw data to state estimation - the most appropriate strategies will be developed to serve best the shared perception ecosystem and targeted application, e.g., operator tracking, mobile robot localization, human-robot interaction, crop type detection etc.

​

Human Robot Interaction for Agricultural Systems - SMART Farming robots and tools are useless if not designed to effectively interact with humans. Therefore research on HRI becomes a key issue in the development of an holistic approach to Precision Agriculture. Communication between a human and a robot may take several forms, but these forms are largely influenced by proximity: Proximate interactions; Remote interaction. Within these general categories, it is useful to distinguish between applications that require mobility, physical manipulation, or social interaction.

Some solutions or research areas include:

  1. Dynamic Autonomy, Mixed-Initiative Interaction, and Dialog;

  2. Telepresence and Information Fusion in Remote Interaction;

  3. Cognitive Modeling;

  4. Team Organizations and Dynamics; and

  5. Interactive Learning.

This task is focused on organizing competences around this topic and creating guidelines and mechanisms for design, development and assessment of Human Robot Interactions in the specific context of SMART farming, and training.

​

Connected Agricultural Machines and Integration of SMART Robots and/or Legacy Equipment with SMART Tools - Connected agricultural machines are at the heart of SMART Farming. Specific standards for electronics and data exchange between different farm machines (e.g. tractor – farm implement) as well as between farm machines and software systems, such as Farm Management Information Systems (FMIS) have been developed. ISOBUS (ISO standard 11783 using ISO-XML as data standard) is an international communication protocol that sets the standard for agriculture electronics. It has become the de-facto standard governing interoperability between tractors and implements from different manufacturers. There is also a clear need to standardize the exchange of more transaction-related data with other actors in the agricultural supply chain. AgGateway (http://aggatewayglobal.net/) is the recognized international organisation that uses the concept of industry cooperation to expand the use of e-business standards and guidelines globally. Hence developing competencies on the existing standards to support connection between agricultural machines and systems is fundamental to develop successful industrial prototypes or systems. Dissemination of these standards among farmers and industrial companies is also an important task and benefit for all stakeholders.

Another fundamental activity here is developing competencies on how to integrate the SMART tools to be developed with both SMART Robots and legacy equipment to pursue plug and play systems support dynamic and smart configuration, adjusted to each context (CAN and ISOBUS - ISO 11783 standards will be used and extended).

​

SMART EQUIP

SMART MONITORING

SMART farming requires SMART monitoring. The primary goal of this line of research is to research and develop multi-modal data collection, merging and integration systems, and support farm operations decisions.

This line of research serves as a perception layer and data source for SMART APPLICATIONS and SMART PRODUCT, and also makes use of the SMART robots from SMART EQUIPMENT.

In this research line, perception is based both on remote sensing, e.g. satellite imagery, weather data, and the uses UAVs for low altitude monitoring, and on local sensing, that relies on a network of in situ sensors in the farm fields.
The WP will also focus on the use of Data Analytics to support technical and
operational decisions and create guidelines for business development.

​

The main tasks for SMART EQUIPMENT are:

 

SMART Monitoring - This task is focused on all farm monitoring activities required to support farm operations decisions. SMART monitoring provides real-time multi-modal perception, allowing effective contextual awareness of the farm, supporting timely decision-making and creation of digital twins of farm objects. Concretely, this task will combine information from local sensors with satellite information to detect patterns and evolution of patterns (e.g. using multi-spectral imagery and fuzzy image fusion algorithms with reinforcement aggregation operators).

This task is organizing competences around this topic and creating guidelines and mechanisms for design, development and assessment of SMART Monitoring in its two facets:

  1. Remote Sensing: Satellites and UAVs - the software necessary to process multi-modal data from both high and low altitude remote sensing. Concretely, weather and land surface products from EUMETSAT and NOAA, radar (Sentinel), hyperspectral or multispectral data taken from UAVs, and in situ sensors. Deep machine learning algorithms and ensemble methods will be used for data classification and data fusion (radar + hyperspectral + weather).

  2. Local Sensing: IoT/Cognitive Cyber Physical - Development of the necessary sensors to support farming decision making, such as low-cost weather stations, vision systems or wireless sensors spread on the field.

​

Data Analytics - The aim of this activity is to provide a big data analytics platform that meets all real-time big-data requirements to gather and feed analytic results back from collected data form the different information sources in a farm environment so real-time predictions for farm management systems can be achieved. Such a platform will automatically implement all the real-time and big-data requirements of the environment defined by an ontological approach. The platform should be concretized to be executed on concrete targeted environment (CPS), and configured for taking into account the several data sources within that environment. The Big Data Analytics Platform will support decisions on reconfiguration, so dynamic run-time decisions can be made during operation. Due to the spatio-temporal nature of the phenomena appropriate interactive and multi-linked data visualizations will be developed to support the human understanding of the main results of data analytics and, in general to support the analyses and the decision making processes.

​

Communications & Positioning Systems - The main goal of this task is to do research and innovation about the telecommunication issues that need to support the interactions between smart equipment and between them and applications. In this activity will be exploited several communication systems working as heterogeneous networks. Cellular communication systems (3G/4G and 5G), Wi-Fi, LoRA and satellite communication will be considered to control UAVs and UVGs. In order to obtain high communication reliability will be studied and implemented diversity techniques. Protocols and algorithms to improve the security of UAVs and UVGs avoiding jamming and spoofing techniques will be implemented. Also, some attention will be considered for positioning systems (GPS, Galileo, GNSS, etc..) using several systems for positioning receivers

SMART MONIT

SMART PRODUCT

This Line of Research aims to maximize the added value and potentiate the National generated horticulture, fruticulture and viticulture products accordingly to the current paradigm for sustainable farming systems and consumers demanding:

  1. Adjusting harvests to market opportunities

    1. Improving of environmental conditioning after harvest,

    2. Increasing nutritional and/or functional added value to intensive fruit farming

  2. Bio control on intensive Smart Farming

    1. Reduce the use of plant protection products aiming to eliminate pesticide residues

    2. Novel compounds responsible for priming biotic stress defense in grape;

  3. Increase Safety and Shelf life

    1. Control of ripening: application of bio-stimulants and biotechnological strategies to increase in shelf life

    2. Increasing knowledge on the physiology of Rocha pear, under the influence of annual variability and postharvest storage conditions, to ascertain the underlying mechanisms that determine fruits quality and develop competitive pre- and postharvest technological solutions to maximize Rocha pear quality

  4. Implement traceability and certification of origin for horticulture, fruticulture products

​

The main tasks for SMART USE are:

​

SMART valorization of sensorial attributes - Mapping physiological parameters for digital markers of fruit ripening. Development of novel bio-cyber platform for intuitive predictive of the sensorial/nutritional characteristics for real-time management of productive process on Rocha pear, Tomato and Grape.

​

SMART Environmental conditioning - Based on the market forecast for demand, timing for harvest is synchronized to market opportunities by improving environmental conditioning.

 

SMART Bio-control - Elucidation of mechanisms responsible for resistance/ susceptibility of grape cultivars, Horti- Fruti- culture against fungal diseases; Use of microbial consortia for priming biotic stress defense, aiming to reduce/eliminate pesticide traits from crops.

 

SMART Shelf life - Identifying internal browning disorders (IBD) predictors, especially under current legislative restrictions concerning synthetic antioxidants (e.g. diphenylamine, DPA) during long-term postharvest storage, including e.g. orchard management practices, soil characteristics and crop nutrition (responses to phosphorus (P) and potassium (K) fertilization).

 

SMART product traceability & certification of Origin - Establishment of low cost real time methodology for product traceability & certification of Origin as well as its integration with the remaining value chain (compatible with ISO 22000/ISO 22005). Innovative analytical methods supported by digital platforms consumer user friendly.

SMART PROD

SMART APPLICATION

The main objective is to develop Smart Decision Systems (SDS) as intelligent computer based applications to support the circular economy, and optimise resources usage and processes.  This work is specifically focused on the devlopment of an ecosystem of applications (apps) to support the decision-making process of the stakeholders in the agri-food sector.

The SDS incorporate heterogeneous sources of data (e.g., sensor data, drone (UAV and UGV) images, weather information, etc.), most of them in real-time, that will be stored and processed (using efficient algorithms for complex analysis) in a cloud environment producing automatic actions in several types of actuators (machine-to-machine actions), e.g., automatic irrigation, fertilizers, phytopharmaceutical products, etc.

The SDS can produce automatic reports, efficient visualization graphics and manual control buttons in intelligent apps for smartphones allowing stakeholders to query historical information, to better understand what happened, but mostly to anticipate when something will happen and how to make it happen (e.g., when to apply a treatment and what quantity and its impact on revenues).

​

The main tasks for SMART USE are:

​

Forecasting Tools - The basic goal behind this task is developing methods and capabilities required to implement the forecasting tools that are needed in a SMART Agriculture environment. The use of Machine learning and Data Analytics developed in the context of SMART MONITORING will be applied to make predictions of the future based on past and present data. This task is focused on organizing competences around this topic and creating guidelines and mechanisms for design, development and assessment of SMART Agriculture forecasting tools:

  1. SMART yield forecasting - Yield forecasting will not only be based on past and present data about yield numbers and by taking snapshots (images) of the fields during crops growing. It is possible to correlate these snapshots with yields but is also possible to predict, for instance fruit yields, giving enough time to producers to book refrigerated warehouses.

  2. SMART Energy and Water consumptions forecasting - Forecasting energy and water consumption is also fundamental to estimate production costs or even booking in advance the necessary supply.

​

Decision Support and Optimisation Tools - The basic goal behind this task is developing methods and capabilities required to implement the tools to help farmers take best decisions in a SMART Agriculture environment. The use of Artificial Intelligence Techniques and such as Machine Learning and Data Analytics will be used to make SMART decision making and optimisation tools for different tasks and activities. This task is focused on organizing competences around this topic and creating guidelines and mechanisms for design, development and assessment of SMART Agriculture optimization tools:

  1. SMART Cover Crop Usage - Implement tools and/or services to optimise cover crops usage by helping farmers deciding where and what cover crop should be planted to manage soil erosion, soil fertility, soil quality, water, weeds, pests, diseases, biodiversity and wildlife.

  2. SMART Crop Rotation - Develop decision support models mainly based  on two concepts, the cropping plan and the crop rotation decisions.

  3. SMART Fertilization and Pest Combat - To develop methods and tools to create SMART or Precise Fertilization technology to help farmers fertilize just the right areas with the right amount.

  4. SMART Irrigation - These intelligent decision tools can use data from remote or local sensing or from local IoT sensors, electromagnetic surveys. Acquiring, spatio-temporal variability of soil hydraulic properties when aiming for precision  management of the resource.

  5. SMART Sustainable Energy Facilities  - Develop methods required to implement decision supporting tools to enable the proper sizing and location of sustainable energetic infrastructures, to enhance energetic supply resilience of the farms and the neighbourhood population.

  6. SMART Crops optimization - The goal of this sub activity is doing all research and innovation required to implement tools and/or services to optimise crop production, including better yields with better quality and safer environment (less water, less fertilizers, less pesticides).

​​

SMART APP
bottom of page