Sensors for Sustainable Digital Agriculture

With global demand for food projected to increase by ca. 50% by mid-century, the agri-food sector must produce this food in an environmentally sustainable manner. This is particularly challenging for an industry that is highly dependent on fossil fuels and where total system losses (waste and inefficiencies) are ca. 50%. It is against this background that Digital Agriculture has a pivotal role to play in enhancing overall operational efficiency. The agri-food sector comprises several stages, a continuum, from on-farm production, through processing, retail, the consumer (citizen) and beyond into an agri-food bioeconomy. Digital Agriculture offers the opportunity to enhance overall operational efficiency of the full agri-food chain, through: (i) gathering data of the quality and scale needed to understand, manage and monitor system performance; (ii) implementing data-driven systems that are designed to optimise (or near optimise) operational efficiency (and hence resource use efficiency); (iii) providing continuous system monitoring, enabling iterative feedback-based system improvement; (iv) providing regulatory overview (monitoring) of compliance with product assurance, environmental impact regulations; (v) providing the consumer (citizen) with credible high quality, verifiable information on the provenance, quality and safety of their food; (vi) providing verification of the provenance and quality of feedstock for a sustainable agri-food based bioeconomy. While digital agriculture has the potential to integrate these features, the critical factor in delivering effective operational systems is the data - its quality, quantity, time-sensitivity, scale and availability (including cost). System variability varies from stage to stage along the agri-food continuum (pre- and post-farm gate), with pre-farm gate (viz. on-farm production) being the most variable as operations are subject to weather, geo-spatial, bio- and market (demand and price) variability. This paper discusses the range of data sources required to achieve enhanced system efficiency, their availability, capture (i.e. sensing technology, scale, costs) and integration into suitable decision support and control systems. It concludes that while several new and existing sensing technologies offer a kaleidoscope of data sources, the real challenges are addressing data quality, variability, scale; and the analyses and integration of these data to deliver effective commercially implementable models delivering enhanced operational performance, monitoring and feedback.