As Australia searches for a technology-based global competitive advantage, ‘agtech’ comes close to the top of almost everyone’s list. That’s easier said than done.
The size of the opportunity – but also of the challenges – for agtech is examined in a recently released report by the Australian Council of Learned Academies (ACOLA). The report looked at a range of technologies, including the Internet of Things, robotics, machine learning, nanotechnology, and biotechnology, which it states may result in novel developments in Australian agriculture. Some examples of how these technologies could change farming practices include:
- a connected ‘swarm’ of contact sensors (e.g. physically placed on livestock), proximal sensors (located nearby but not on the animal) and remote sensors (e.g. airborne or satellite) can create sensor output that is more certain, accurate and dependable than ever before. A farmer could be able to locate his or her cattle, for example, at any time across widespread farming lands. This in turn could lead to innovative land use that could change the notion of a ‘farm’ (for indigenous farming businesses using more sustainable practices across large tracks of native title lands);
- machine learning systems could be used to improve, for example, irrigation systems by estimating daily, weekly or monthly evapotranspiration from meteorological station temperature data, or to improve animal welfare using data collected from drone images or mobile surveillance to identify signs of stress and examine movement patterns, changes in body weight etc. to predict whether the animal should be culled.
The report identifies that a clear national approach to the use of agricultural data will need to underpin many of these technologies. Part of that will need to be resolving outstanding legal and policy issues surrounding the use of agricultural data, including:
- power imbalances: there are imbalances in bargaining power between farmers and the large companies that produce data-dependent technology and therefore are in control of the aggregation and sharing of that data. For example, many of these companies use standard-form contracts which farmers do not have an opportunity to negotiate.
- transparency: farmers have privacy, surveillance and data ownership concerns that can prevent them from trusting technology businesses, researchers and governments will handle data with integrity.
- reducing burdens: some areas of the law need clarification to make sure that the bureaucratic burden on a new technology does not negate any potential benefits. For example, it is unclear to what extent some agricultural data, such as GPS location data, would be considered personal data under the Privacy Act 1988 (Cth), in which case a set of privacy principles would apply to its collection and dissemination.
- clarifying consequences of errors and misuse: what happens when faulty data analysis (e.g. based on an incomplete data set) leads to poor decisions and economic loss? What happens when technology providers inappropriately share data with third parties? The answer will be complex and fact-dependent, and potentially cut across varied fields of law such as tort, occupational health and safety, contract and consumer protection.
Start with the farmers
The Report states explicitly that ‘[f]armers should be active participants in all discussions and decisions in this domain’ (being the creation of a national approach to the use of agricultural data), and acknowledges the need for a farmer-centric approach to digital agricultural systems development that is sensitive to the diversity in farmers’ aspirations, capacities, enterprise mix, farm size, location and climate.
The Report also, somewhat coyly, says that ‘[a]ttitudes to technology and its adoption by primary producers are complex and multifactorial’ and recognises that Agtech initiatives need to be built around certain data principles, including the following:
- making it easier for farmers to collect data;
- avoiding overcomplicating matters;
- minimising the steps between data collection and useful knowledge;
- focusing on helping farmers to test and improve their own knowledge rather than replacing it;
- recognising that digital systems and hard data are not the only sources of information farmers use to make decisions; and
- recognising the utility of ‘low powered communication systems’, as high speed connectivity is important for real-time data-based decision-making, but not for all technologies.
But the challenge is not about farmers being Luddites. Farmers are already required to collect an enormous amount of data just to meet increasingly burdensome regulatory requirements, and are feeling increasingly overwhelmed by these obligations (to which we can testify as one of us is an amateur farmer). The various data collection systems are clunky and non-intuitive: case in point in the National Livestock Identification System into which farmers enter radio-frequency identification (RFID) ear tags to allow tracing of cattle from farm to abattoir. Separate non-interoperable data collection systems apply between different agricultural sectors (e.g. cattle vs fruit) and whether the farmer is involved just in production and growing or is integrated into processing. On top of this, large private buyers, such as wholesalers and supermarkets, are now requiring farmers to participate in their data collection and reporting systems.
It is difficult for farmers to see the value in data collection – what they see is cost and compliance risk in collecting data into incomplete and unwieldy databases.
The way forward
The most telling conclusion in the Report is that farmers ‘need a clear value proposition in order to be willing to adopt the new technologies’.
Farmers need to see data as an opportunity rather than a burden, and therefore be willing to take up new technologies that are data-dependent. The Report suggests:
- a voluntary agricultural data code of practice to improve clarity regarding data licence terms, and to regulate the ownership, sharing, privacy and security of data;
- a new legal framework for data sharing and a consumer data right for farmers when they deal with agtech data companies;
- a binding industry code or standard (e.g. under the Competition and Consumer Act 2010) to protect farmers’ ownership of the data collected by agtech companies; and/or
- improving digital and data literacy among farmers to better enable them to navigate the new technological landscape and advocate for themselves. The Report suggests that massive open online courses (MOOCs), Vocational Education and Training (VET) or TAFE could be appropriate means to teach digital literacy skills.
Australia’s success in agtech first requires some hard spade work on the data inputs