Smart, Sustainable and Safe Food Systems – a roundup from the SFN+ 5th Annual Conference 2023

​Configuring our food systems to meet global nutritional needs will require a much broader focus than farming and agriculture. With the challenge of climate change and pressure on environmental resources, innovation is needed at a systems-level, all the way from field to fork. As the SFN’s 2023 Annual Conference showed, projects funded by the network are targeting every intervention point to deliver ‘Smart, Sustainable and Safe Food Systems.’

Enabling Net Zero Food & Drink – the need for End to End innovation by Professor Ian Noble, VP R&D – Research & Analytical, Mondelez International
The global scale and multifaceted nature of food systems makes feeding the human race a “mind-bendingly complex” problem that can’t be fixed using simple, single-point interventions. But this creates opportunities to introduce transformational change that ripples outwards across different levels. Achieving this, however, requires systems-thinking that goes beyond seeing food chains as linear processes. Instead, we need to adopt ‘design-led’ approaches that explore problems with creativity, whilst keeping stakeholder needs front and centre. Trying to introduce even an apparently simple change – such as reducing packaging on chocolate Easter eggs – will only be successful if producers still have a sustainable business and customers accept it.
This approach might sound daunting, but it also creates a space for conversations that throw the door open to new possibilities of doing things – and for collaboration across different disciplines. As Ian said, “If we approach the problems through a design mindset, this opens up new areas of creativity. Furthermore, food is such a broad area that there is space for everyone. Whatever your discipline – whether it is artificial intelligence or crop science – there is room for you to engage. And with the scale of the transformation that has to happen across the board, we will need everyone to be part of it.”

The featured projects:

Urban Cultivate: Integrating environmental, spatial, and social data to assign food supply chain functions across urban space by Dr Daniel Evans, Cranfield University
Urban Cultivate is exploring whether under-utilised vacant spaces within cities can be converted into growing centres for self-sufficient, local food systems. In a pilot study based in Islington, London, they used satellite images to identify 100 underused sites around Tufnell Park Road, then collected data on a cluster of these – including air and soil quality, microclimate, urban connectivity and connection to utilities. Following this, the team applied STFC Data Science Capabilities to design an optimization model which could calculate each site’s suitability value for different food-growing functions – from cultivation and composting right through to food exchange – which is then presented in an accessible, public dashboard.

Increasing the Nutrient-Use Efficiency and Crop Productivity of Hydroponics using SMART Sensors and 3D-Multispectral Crop Imaging Project PI: Professor Chungui Lu, Nottingham Trent University. Presented by Dr Gadelhag Mohmed, Nottingham Trent University
Could artificial intelligence (AI) and machine learning reveal ‘ideal plant growth recipes’ hidden in large datasets, and help make vertical farming systems more resource efficient? To investigate, this project generated Big Data from indoor growing systems by measuring environmental variables using a range of sensors, and combining this with output data such as shoot weight and leaf area. Using the STFC Hartree Centre, they constructed a cloud computing and data processing platform to develop a neural network AI model which could predict optimum growing recipes. These recipes are now being validated, and the team hope to refine the model with tailored algorithms for specific crops.
Find out more in this SFN+ blog post

Circular urban vertical farming. Data, models and optimisation of waste flows by Professor Peter Ball, University of York
Municipal waste could be a significant resource for circular, sustainable urban agriculture but without the proper data, these flows can’t be captured. Working with a vertical farm in York, this project developed two digital demonstrators to model how waste flows from urban sources could be optimised to provide a) feed nutrients for crops and b) construction materials. With support from IntelliDigest (see below), the team analysed the simulated data to see how closely the output matched the nutrient requirements of growing crops. The team also worked with a brewery and WASWARE to create biocomposite materials using grain waste from York beer production to demonstrate the proof of concept.
Find out more in this SFN+ blog post

Building smart urban farming data systems: A case study by Professor Catur Sugiyanto, Universitas Gadjah Mada, Indonesia
By combining digital technologies and geographic information systems with farmer surveys, this project took a deep dive into exploring whether urban farming can improve food security for vulnerable urban dwellers in Southern Asia, focusing on Yogyakarta in Indonesia. Their findings revealed that Yogyakarta has a diverse urban farming system, with some areas of the city specializing in specific products – such as fish in the south west regions. The surveyed families farmed not only for self-consumption but also to sell produce for additional income. Notably, a significant proportion of urban farm managers are women, enabling them to become more active in the economy.
“We found evidence that urban farms really can increase the incomes of families in Southern Asia. For our case study of Yogyakarta, household incomes of urban farmers were 50% over the regional minimum wage.” Professor Catur Sugiyanto, Universitas Gadjah Mada,

Spotting the Rust: Tracking disease progression by static and aerial imagery by Dr Melina Zempila, Science and Technology Facilities Council
This proof-of-concept project aims to develop a hyperspectral imaging-based classification system to enable early detection of a devastating wheat disease, yellow rust fungus. To inform the model, the project team combined hyperspectral and multispectral imaging data from laboratory-grown plants at different stages of rust development with aerial images of healthy and diseased wheat crops. Both approaches revealed that the reflectivity of wheat plants over certain wavelengths markedly increases during infection with yellow rust. From this, the team developed classification algorithms and differential approaches to identify diseased plants from novel data.

Novel process for extracting nutrients in food waste for sustainable and resilient urban farming and food packaging by Dr Ifeyinwa Rita Kanu, IntelliDigest
IntelliDigest has ambitious plans to catalyse the adoption of circular food systems, particularly through recovering bionutrients from waste streams using bioreactors containing natural digestive enzymes. But this will require real-time information on food waste composition, to optimise the combination of enzymes used. With STFC’s Central Laser Facility, IntelliDigest have explored a range of spectroscopic techniques (including near- and mid-infrared, and Raman Spectroscopy) that may be suitable for developing into a photonic sensor. Using the STFC Dafni platform, they have also developed a pilot interface which tracks the economic and environmental impact of diverting different food waste streams.
Find out more in the SFN+ blog post.
“Pilot growing studies indicate that our biodigest product has high macro- and micro-nutrient content for food production – and all of this is based on harnessing the power of natural enzymatic processes.” Dr Ifeyinwa Rita Kanu, IntelliDigest

SFN+ Strategic Partnerships Project: Digital Twin for FSA inspection process by Professor Sonal Choudhury, University of York
Following an SFN+ sandpit event, this consortium was launched to explore how advanced technology interventions – particularly digital twins – could help upgrade Food Standards Agency (FSA) meat processing and inspection operations from the 18th to the 21st century. Using discrete-event simulation (DES) modelling (which simulates processes where events occur at specific instances in time), the team produced a model of a pig abattoir process flow, enabling different scenarios to be tested. In particular, the model enables virtual testing of technology, people, and process configurations to explore ‘what-if’ scenarios to predict and optimise the system behaviour.

​You can watch the recorded session here.