A study recently published in the journal Sensors comprehensively explored how artificial intelligence (AI)-powered data infrastructures could enhance sustainability in the agri-food sector. Focusing on soft-fruit production and brewery operations, it highlighted both the potential benefits and key challenges associated with these technologies.
Study: Embedding AI-Enabled Data Infrastructures for Sustainability in Agri-Food: Soft-Fruit and Brewery Use Case Perspectives. Image Credit: MONOPOLY919/Shutterstock.com
The researchers emphasized that while AI can optimize resource use, enhance food quality, and lower carbon footprints, successful implementation depends on addressing critical issues such as data quality, user engagement, and model generalizability.
Agri-Food Sector's Transition Towards Net Zero
The global food system contributes approximately one-third of human-caused greenhouse gas (GHG) emissions. This necessitates a rapid transition toward net-zero emissions within the agri-food sector. In the UK alone, the food and drink industry plays a significant economic role, employing around 430,000 people and contributing approximately GBP 35 billion to the national Gross Value Added (GVA) annually.
Achieving net-zero goals requires substantial changes across the entire supply chain, from production to delivery. Recent advancements in data collection and monitoring have introduced new methods to improve environmental management within supply chains.
Transparency is crucial for identifying risks, tracking progress, and addressing sustainability challenges. However, existing tools such as sustainability scorecards, self-disclosure systems, and carbon calculators often lack accuracy, are data-intensive, or require specialized skills. Optimizing food production to maximize yield while minimizing resource use and environmental damage is essential. Emerging technologies like AI and the Internet of Things (IoT) have the potential to achieve these goals, but further research is needed to understand their capabilities and address related challenges.
Using AI and IoT in Agri-Food Sustainability
This paper investigated the feasibility and challenges of using AI and IoT technologies to reduce carbon emissions in the agri-food sector. It focused on two use cases: strawberry production and brewery operations. The authors collaborated with a UK-based strawberry producer and several breweries to test IoT-enabled platforms in real-world settings.
In the strawberry case study, IoT sensors were employed to monitor environmental factors inside and outside polytunnels, including temperature, humidity, light, and soil moisture. Then, AI algorithms analyzed this data to optimize growing conditions and predict production yields, thereby enhancing resource efficiency and food quality.
Similarly, in the brewery case study, sensors were used to track parameters like fermentation temperature, pressure, and dissolved oxygen. This information improved efficiency, enhanced product quality, and facilitated accurate carbon footprint calculations for brewing processes.
The researchers followed a user-centered design approach, involving stakeholders throughout the process to ensure the usability of the developed solutions. This included organizing workshops and sessions with soft-fruit growers and brewery stakeholders to identify opportunities and challenges associated with AI-based solutions.
Challenges and Insights from Real-world Deployments
The study identified several key challenges in deploying AI-based solutions for sustainability in the agri-food sector. User engagement emerged as a critical factor, with the researchers emphasizing the need to involve stakeholders early to ensure the practicality and acceptance of new technologies.
Data quality posed significant issues, arising from environmental variability, sensor reliability, and inconsistencies across different production facilities and farms. The generalizability of AI models was another challenge, as models trained on one farm often performed poorly when applied to others. This was particularly evident in the strawberry case study, where variations in soil types, polytunnel configurations, and plant varieties affected model accuracy.
Socio-technical barriers included high capital costs, a lack of standards and regulations, and the necessity for domain-specific expertise to enhance model performance. Economic feasibility was a particular concern for small and medium-sized enterprises (SMEs), which dominate the agri-food sector.
The authors also addressed ethical concerns regarding data ownership and privacy. They suggested approaches like federated learning to safeguard sensitive data while allowing for collaborative model training. Additionally, they highlighted that optimizing resources and processes could improve food quality and consistency.
Practical Applications
This research has significant implications for the agri-food sector. AI-enabled decision support systems can help producers optimize resource use, reduce carbon emissions, and improve food quality. IoT sensors for real-time monitoring can provide insights to boost efficiency and identify areas for improvement. Transparent carbon footprint calculators can assist businesses in tracking and reducing their environmental impact.
Conclusion and Future Directions
In summary, deploying AI-based data systems for sustainability in the agri-food sector faces challenges like user engagement, data quality, model generalization, and socio-technical barriers. The authors emphasized the importance of user-centered design, semantic data integration, and transparent carbon footprint calculators. It also highlighted the need for a comprehensive approach to connect AI models, data availability, and emission calculations.
Future work should address these challenges by developing advanced AI applications such as semantic data integration, synthetic data generation, and multi-objective optimization systems. Promoting user engagement, increasing awareness of AI and sensor technologies, and bridging the knowledge gap between researchers and industry professionals would further support these advancements.
Journal Reference
Markovic, M.; &.; et al. Embedding AI-Enabled Data Infrastructures for Sustainability in Agri-Food: Soft-Fruit and Brewery Use Case Perspectives. Sensors 2024, 24, 7327. DOI: 10.3390/s24227327, https://www.mdpi.com/1424-8220/24/22/7327
Disclaimer: The views expressed here are those of the author expressed in their private capacity and do not necessarily represent the views of AZoM.com Limited T/A AZoNetwork the owner and operator of this website. This disclaimer forms part of the Terms and conditions of use of this website.