Happy new year. For the past 5 years the start of the new year for me has been accompanied by Radio 4 updates on the annual Oxford Farming Conference (http://www.ofc.org.uk/) and the Oxford Real Farming Conference (http://www.oxfordrealfarmingconference.org/). My listening was always laced with envy at the thought of all the UK farmers getting together to get their thoughts straight prior to the year ahead. Perhaps we should have the BES annual meeting in January too.
I bring it up however because of the “Rethinking Agricultural Systems in the UK” event we helped to run just prior to Christmas (http://futureagriculture.wordpress.com/). There around 100 attendees presented and discussed where UK agriculture is going or needs to go and I was there to gain inspiration about what might be new computational solutions. I went into it very aware that many farmers are often very keen to try out new approaches. Indeed, one of the lasting memories from my childhood in rural Aberdeenshire in the 1980s was just how experimental farmers were. It was with some surprise therefore that I read the UK agri tech strategy a couple of months ago (https://www.gov.uk/government/publications/uk-agricultural-technologies-strategy) where they say
“This is the first time the UK Government, science base and food and farming industry have come together to identify and develop the opportunities and strengths of the UK agricultural technologies (agri-tech) sector as a whole.”
“Oh”, I thought, “Talk about late in the day”. But I see the opportunity. Apparently it is only now that the UK government has started working with its science base and the industry to investigate some major upgrades to their agricultural operating system (someone will tell me soon this is a gross caricature). Over the subsequent months, after much conferencing, I’m tempted to adopt the opinion that we’re not even scratching the surface yet with how we might revolutionise UK agriculture with technology.
Back to the Rethinking Agricultural Systems in the UK workshop. We had a tremendous variety of ideas and proposed solutions and at the end of the last day we had a mini workshop to talk through the different computational ideas that came out of it. So here is a shortlist of the main items we were excited about.
- Ubiquitous sensors and devices. This one’s obvious. It is now less an issue of whether there exists technology to measure something and more simply “what do you want to monitor?” We can now deploy nutrient sensors, phenology monitors, robotic weeders, you name it. There was particular excitement about ubiquitous hyperspectral technology – enabling us to monitor multiple aspects of the agri-ecosystems with static or machine (e.g. tractor, drone) mounted cameras. Finding innovative ways to see into the soil (e.g. seeding in some “monitor” plant species that act as leading indicators of impending crop problems) was seen as a potential valuable area for future research. Enabling the biodiversity in a habitat to be quickly quantified through pictures from a camera phone to help farmers quickly assess how well they qualify for various biodiversity schemes was also excitedly discussed.
- Real time, multi-scale, cost benefit analyses of farming options. This continually came across at the workshop – that while farmers are individually making rational decisions based on their own particular cost-benefit analysis (based on multiple criteria like profit, effort, lifestyle, experience, ambition) there are no tools available, or few analyses being conducted, that allows their decisions to be set within the context of what options are available, and what is being decided at regional, national and international scales. This could be a profitable area for research but here is somewhere that the technology is letting us down – it can be extremely challenging even for tech-experts to conduct analyses that combine detailed local information such as spatial-time series of how land is being managed with much larger scale information such as the degree to which the multiple goals of the UK land management strategy are being achieved (in terms of yields, protected areas, clean waterways). We also need such analyses to be able to assess the risk of losses (or gains) from different events. What are the major risks threatening crop productivity? E.g. would a system allowing multiple cost-benefit analyses allow us to predict that the UK farm herd was susceptible to a foot-and-mouth outbreak? It is this sort of “systemic risk” that we need new systems for multi-scale analyses to be able to predict. What about the risks from extreme weather? New pest strikes? Would farmers be tempted to try more diverse crops if they could see some metric of the risk of different amounts of financial loss from different planting portfolios?
- At a simpler level, there was excitement by the realisation that there is little out there reporting the detailed resource input output network of farms. OK, we’re aware that animal movements between farms are now tracked but what about where fertiliser comes from and where nutrients go to? We saw presentations earlier in the day showing the spatial dynamics of nutrients into and out of farms, and highlighting that far more efficient nutrient import-export relationships might be found, where farmers source their phosphorous (for example) more locally and more cheaply. Someone even mentioned the possibility of a nutrient-Ebay for farmers. Before all that though – we need better ways to record and track these resource flows. (in terms of monitoring agricultural systems as systems, this bit is like keeping a monitor on the blood flow system of a living organism)
- Enhancing the existing knowledge transfer network amongst farmers was also discussed. We began by reminding ourselves that farmers already have very good knowledge transfer networks. However, what might help – e.g. would it be valuable to provide farmers with the opportunity to receive live updates of what farmers in similar circumstances to them are doing right now and why (e.g. why are you now planting x?). With more farm-level information, multi-scale cost benefit analyses and resource transfer information (items above) such technology might allow farmers to barter about how they use their land to achieve multiple goals (e.g. producing a certain tonnage of oilseed rape while setting aside a certain proportion across a region for experimental crops or growing trees). Highlighting a pest-strike in a particular area, putting farmers in touch with agronomists.
For me this was a brilliant way to end the working year and, looking back on it now, I find I probably don’t need the farming conferences to see some useful new things to study in 2014 to see if I can have a positive effect on UK agriculture. Best wishes for 2014.