Growing food sustainably and locally brings benefits to people and the planet. A Research Associate at the Alan Turing Institute, Dr Rebecca Ward presents a collaborative live project that places data and physics front and centre to help an underground urban farm grow smart.
Cities are growing. According to the United Nations, by 2050 two out of every three people are likely to be living in cities or urban centres. Increases in populations bring extra demands on resources and services in urban areas, putting cities under pressure to provide what people need – including a constant and sustainable supply of food.
Growing and delivering food brings an associated carbon cost – particularly in the transportation of goods. Exploring new ways to feed a city that minimises our carbon footprint and ensures energy use is kept to a minimum is of increasing importance.
The Energy Efficient Cities initiative (EECi) at the University of Cambridge Department of Engineering has at its heart the ambition to understand how we can reduce our energy demand and environmental impact in towns and cities. With the Alan Turing Institute and CSIC, we have been developing smart and sustainable approaches to help improve the efficiency, productivity and profitability of urban farming in collaboration with the world’s first underground farm, Growing Underground, which is located 33 metres below the pavements of London’s Clapham High Street in former World War II air raid shelters.
The physics model is helpful as it enables us to test and evaluate environment scenarios in ways that are not possible in real time because the farm is a commercial entity and operations cannot be interrupted. We can play with the physics model and find out if there are benefits to be gained, for example, from turning the LED lights on two hours earlier as temperatures dip, or if is it better for the plants to wait until it becomes cold before switching the lights on. Dr Rebecca Ward, Research Associate
The Growing Underground farm is operated by Zero Carbon Farms Ltd. and produces microgreens – pea shoots, basil, coriander, parsley, salad rocket, pink radish and mustard plants – from an area roughly the size of a tennis court, hydroponically (without soil) using carpet cut-offs destined for landfill. The company packs on site and takes the harvest to market less than a mile away for distribution across the capital, keeping food miles, pollution and food waste to a minimum. This method of farming is space-efficient, having a smaller footprint than a conventional greenhouse and uses 70% less water than conventional agriculture methods. The primary cost is the LED lights, but as these provide both lighting and heat, no additional heating is required and plants can be grown year-round. By using 100% renewable energy, the carbon footprint is kept to a minimum.
Our research team has worked with the commercial concern for the past seven years to help it realise its goal of becoming a zero-carbon food company. From the start the owners recognised that data would play an important part in future success and researchers began by installing sensors throughout the 528m² area to collect data about everything contributing to the operation of the farm and its yield.
The data offers real-time information and a virtual representation – a digital twin – of the farm which my colleague Melanie Jans Singh and I have created together with software engineers from the Alan Turing institute helps us to monitor, learn, feedback and forecast information to make the real-life twin work better. This information enables the owners of Growing Underground to understand how different elements of the tunnel environment affect the crop yield, and to track and fine-tune variables – lights, water, heat, nutrients, airflow and humidity – to optimise yield while conserving energy. The value of data here is evident – Growing Underground has already reduced the time it takes to grow some crops by 50% and all crops by an average of 7%, and increased yields by 24%[1].
Alongside the development of a digital twin, the physics of the farm was analysed. This is an important addition because, while the digital twin shows real time events, we also want the farm to be able to respond to other variables such as the weather outside the tunnel environment. The physics model is helpful as it enables us to test and evaluate environment scenarios in ways that are not possible in real time because the farm is a commercial entity and operations cannot be interrupted. We can play with the physics model and find out if there are benefits to be gained, for example, from turning the LED lights on two hours earlier as temperatures dip, or if is it better for the plants to wait until it becomes cold before switching the lights on. The model can also help with variables that are difficult to measure such as the antiquated ventilation system that has basic controls and gives little indication of the impact of changing the dial settings.
We can use the physics model to infer what the effects of the ventilation might be based on what the data are telling us. Working with the Alan Turing Institute, we have also developed a continuous calibration process specifically for the digital twin that brings in the sensor monitoring data and continuously calibrates the physics model. It is inferring values for two important parameters for the model – the ventilation rate and the internal air speed in the tunnel – which are important because the internal air speed governs the convection rate (the heat loss from the plants to the internal air) and the ventilation rate governs the heat loss from the internal air to the outside of the tunnel, and vice versa. By running this calibration process, we're generating a continuously calibrated model that at any moment in time is the best representation of the physical system that it can be.
The digital twin, or Crop Research Observation Platform (CROP) as it is known, incorporates a visualisation platform that allows operators to look inside the farm and click on any of the sensors and see what the values are in real time. It gives an immediate visual view of the state of the farm and, on zooming out, shows where that farm is located within the wider urban environment. This is going to be very useful as the farm expands into the tunnels beyond its current location and we will be able very easily to extend the visualisation and incorporate new sensors as they come on-line. This visualisation has been enhanced over the past 12 months to include recent historical data and present information about cumulative performance that enables the farmers to estimate whether conditions have been met to grow the required amount of micro greens to fulfil orders. We plan to incorporate crop yield data next and interpret the crop yield in relation to the environmental conditions in order to provide a visual interpretation of the crop performance.
While micro greens are not going to feed a hungry city, the Growing Underground project shines a light on potential mitigations to food security – as the technology behind LEDs, wireless sensors, cloud computing and Internet of Things progresses, and if cheap renewable energy and battery storage becomes available, urban farming and growing food in an array of different buildings becomes increasingly commercially feasible. Growing plants in an urban setting has benefits beyond providing sustenance. Encouraging plant growth for food production in our schools can help children to better understand where their food comes from and including communal growing spaces such as allotments or gardens in urban re-generation projects can have positive physical and mental health benefits too.
Putting data front and centre of this business has delivered more information for better decision-making. This is of value to Zero Carbon Farms as they expand the Growing Underground farm and explore different crop types but it also points to the potential of different ‘greenhouse’ spaces in cities – discarded, empty or re-purposed buildings – to grow the food we need sustainably and locally.
Quick Lnks
Growing Underground – how smart monitoring is helping an urban farm to flourish
https://www.cam.ac.uk/stories/growingunderground