The new EU common agricultural policy (CAP) [1] has defined the policies for the climate, the environment and animal welfare, which will apply for the period 2023-2027.To address societal needs about foods and health, the CAP strategic plan [2] supports actions towards the climate transition, sustainable production, animal welfare and environment. Specifically, the CAP promotes sustainable agroforestry practices, such as: biological agriculture, integrated production, agro-ecology, agroforestry and precision farming.Additionally, according to the sustainable development goals of the Agenda 2030 [3], the EU Council also promotes the activities devoted to the general improvement of the animal welfare and to fight the antimicrobial resistance.Precision livestock farming (PLF) is defined as the individual animal management by continuous real-time monitoring of health, welfare, production/reproduction, and environmental impact [4,5,6]. The use of sensors to collect data on animals’ behavior and livestock farming production in PLF has several potentials, including: i) the early detection of diseases and other animal welfare issues; ii) the improvement of production performances; iii) the optimization of available resources usages; iv) the minimization of environmental impact; and v) the increase of livestock farming societal acceptance.Moreover, the development of new technological tools (e.G., RFID, ruminal bolus, activometers, thermographic cameras) has provided new and significant opportunities to collect fine phenotypes and improve welfare, productivity and sustainability in the livestock sector.As a result, it is evident that PLF has a great potential to facilitate sustainable production.The proposed research project is aimed at applying artificial intelligence (AI) techniques and methodologies to data collected in real PLF scenarios to experimentally support the achievement of PLF objectives and potentials. Specifically, our research goal is to find the tradeoff between livestock productivity and animal welfare, while (among others): i) reducing the usage of antibiotics on animals and in food production; ii) improving the food quality (security and safety); iii) optimizing the use of resources; and Iv) favoring the adaptation and mitigation to climate change.Historical and new data collected in PLF systems will be analyzed by means of AI techniques and methodologies, with the aim of developing data analysis and prediction models able to: i) determine the best balance between animal nutrition, emission of climate-altering substances and sustainable production requirements; ii) optimize the ecosystem services provided by sustainable intensive livestock farming; iii) identify animal population most resilient to climate change to be involved in an intra-breeding campaign for genetic improvement. In this field of research, one of the most difficult challenges is related to the need to combine and analyze heterogeneous data coming from different sensor systems.We plan to combine and process existing historical time series of precision milk data recording (milking ability: speed and yield), animal activity (movement, feeding behavior, rumination time), precision NIR, enteric methane emissions and environmental recording collected from a population of experimental subjects.In this research, other phenotypes may also be leveraged, including data on the quality of milk and derived products, such as milk coagulation properties (r, K20, a30, Dairy Aptitude Index), spectroscopic profiles, somatic cell values, temperatures on qualifying points of the animal (i.E. Udder, ocular orbit, abdomen, etc.), biochemical important parameters in the animal welfare assessment, such as urea, beta-Hydroxybutyrate, acetone, etc.This research project will be articulated along the following research lines:Design and learning of predictive AI models to be applied on heterogeneous data collected in PLF scenarios to support decisions on actions to take towards the mitigation of climate-altering substances emission while preserving the animal welfare and production performances.Use of statistical and machine learning features analysis techniques to identify the correlation and tradeoff between the dietary animal administration and the emission of climate-altering substances and impact on production performances.Conducting experimental research aiming at boosting the performances of the learned models.Applying effective methodologies for the reconstruction of missing data on time series collected by means of digital agriculture technologies.Design and development of software systems to support the sensor monitor activities.Bibliography[1] EU Common agricultural policy 2023-2027 (CAP) https://www.Consilium.Europa.Eu/en/policies/cap-introduction/cap-future-2020-common-agricultural-policy-2023-2027/[2] Regulation (EU) 2021/2115 of the European Parliament and of the Council of 2 December 2021(CAP Strategic Plans), https://eur-lex.Europa.Eu/eli/reg/2021/2115/oj[3] Sustainable Development Goals - United Nations . Https://www.Undp.Org/sustainable-development-goals[4] Berckmans D. 2014. Precision livestock farming technologies for welfare management in intensive livestock systems. Rev Sci Tech, 33(1), 189-196. [5] Lovarelli D, Bacenetti J, Guarino M. 2020. A review on dairy cattle farming: Is precision livestock farming the compromise for an environmental, economic and social sustainable production? J Clean Prod, 262, 121409. [6] David Meo Zilio, Roberto Steri, Miriam Iacurto, Gennaro Catillo, Vittoria Barile, Antonella Chiariotti, Francesco Cenci , Maria Chiara La Mantia and Luca Buttazzoni. 2022. Precision Livestock Farming For Mediterranean Water: Some Applications And Opportunities From The Agridigit Project. In Book: Safety, Health and Welfare in Agriculture and Agro-food Systems. DOI: 10.1007/978-3-030-98092-4_5.

- Knowledge of machine learning and deep learning algorithms and techniques for the supervised and unsupervised learning of predictive models on heterogeneous, sparse and noisy data, including data in the form of time series. - Programming skills with python and AI suites.- Experience with DBMS, SQL, NoSQL and Web programming. - Propensity to team working and interdisciplinary research. The preferred candidate should also have knowledge of PLF methodologies and technologies.

Research at UnitelmaSapienza (a young online & distance learning University directly linked to Sapienza University of Rome) is carried out in various laboratories, research centers by various research groups.The Intelligent Information Mining research group (IIM - http://iim.Di.Uniroma1.It) IIM involves researchers from UnitelmaSapienza and Sapienza University of Rome which collaborate in investigating research topics in the following research areas: Machine Learning and Data Mining, Knowledge Based Systems, E-Health and Network medicine, Social Media Analysis and Recommender Systems, Human-Computer Interaction and Web Engineering, E-Learning and Educational Data Mining.The proposed research will be supervised by - Prof. Eng. Damiano Distante, PhD (SSD INF/01, ERC: PE6_7, PE6_10, PE6_11), Department of Law and Economics, University of Rome UnitelmaSapienza, Italy, Advisory member of the IIM research group (https://iim.Di.Uniroma1.It)In collaboration with: - Prof. Stefano Faralli, PhD, Computer Science Department, Sapienza University of Rome, Italy (s.S.D.:INF/01, ERC: PE6_7, PE6_10, PE6_11), member of the IIM research group;- Dr. Miriam Iacurto, (SSD AGR/19, ERC: LS9_3), Dr. Roberto Steri (SSD AGR/17, ERC: LS9_3) and Dr. David Meo Zilio (SSD AGR/17, ERC: LS9_3), Consiglio per la ricerca in agricoltura e l'analisi dell'economia agraria (CREA), Italy (agreement to be established between UnitelmaSapienza and CREA - Centro Zootecnia e Acquacoltura - ZA).