Simone, can I just congratulate you on the way you have brought the themes of this broadranging discussion together - well done

I agree that capacity building is the key to building trust in data-driven agriculture. There are many good examples  of how capacity building can be delivered on farm. We have started developing on on-line digital toolbox for farmers in Australia so that they can access checklists for data licences, information sheets about data control and access etcand the plan is to build this and extend the range of materials and tools to help farmers gain the skills to engage in the dialgoue around the sharing of benefits derived from on farm data collection and aggregation.

I agree that Indigneous and traditional knowledge and farming practices should have special exemptions under any proposed scheme (along the lines of the CGIAR approach of open as possible but closed as necessary which allows for restrictions)

This provides a very interesting example of one way to go forward with this broad and ambitious topic. From my experience, we untertook a large research project that examined  a number of case studies of 17 different Australian agricultural industries (from dairy, fish, forestry grains etc)  and discovered there were very different attitudes in each industry to what data they needed to access and what data they wanted to share and how best to go forward. Issues of connectivity were also included as one of the potential barriers to adoption. Like your workpage, we have drawn some general recommendations for a national approach for an agricultural data policy which each industry can then adapt for their own situations. For anyone interested in the survey results or of any of the findings, they are available here:  https://www.crdc.com.au/precision-to-decision While many different agriculural industries are prepared to share common data, there are some sensitivites with certain crops in terms of biosecurity and environmental compliance where industries are less willing to have that data shared.

I do find a paralell which the notion of accessing genetic resources - while these, like raw data, cannot be owned, the Plant Genetic Treaty does provide an international framework for access and benefit sharing which provides Model contracts for countries to use. Benefits come back to the countries and the communities from where the resources were sourced.While there are obvious differences with data driven farming, on one level a broad principled based approach based on notions of transparency fairness equity would be one step forward in providing a long term solution for our farmers

The EU Code of conduct  as discussed provides a useful example of  how developing 'best practice'  in data sharing can help to build trust in data-driven industry led initiatives.  There are some similarities that can be drawn here to the American Farm Bureau's Privacy and Security Principles of Farm Data that was discussed earlier on Day 1. Those principles operate from the basis that farmers own their data.  It is important to remind ourselves that (in the absence of specific legisated data rights)  there is no 'ownership' of raw farm data until there has been some value adding ie aggregation.  However, the value of such principles and codes is that they raise the awareness of what good data contracting may and should look like eg rights of portability, accessibility and consent etc.   So reinforcing the fact that farmers should benefit from data sharing, voluntary codes such as this can help both farmers and agri-businesses  to ensure that data sharing arrangements are more transparent and equitable. Often in conjunction with such codes, are certification or authetification processes so that those service providers or 3rd parties who comply with the code can signal that fact to their data contributors  and this then gives choice to data contributors and this in turn this builds trust.  Data sharing tools developed in light of such codes also build trust by ensuring that data is interoperable, machine readible and thus easier to share. Facilitating the sharing of data is the key but it is important to build confidence in famers by building trust in the data sharing arrangements  and this will need to come a range of initatives such as farming co-operatives, educative tools and industry codes. 

Some of the  research that has been done on attitudes of farmers to the sharing of their farm data shows a willingness of many farmers to share different types of data with different stakeholders. For example, many are willing to share soil, weather and nutrient/chemical data with Governments on the understanding that this will then be made available to other farmers and agronomists who will return benefits of aggregated services to farming communities. With developing countries, it is these public data sets that need to be available  to farminig communities with as little or no cost. Other sorts of data such as production/yield etc may be shared by farmers and their communities to other trusted communities where the benefits gained from the data and knowledge shared can also be equally distributed among the data contributors. There will always be some data or information, particularly in Indigneous farming communities that may be shared but on a more restricted arrangements (as envisaged by the CGIAR approach to data sharing). It is important that the small holder farmers are given opportunities to learn about the potential value of contributing data while at the same time being given some tools to understand the potential risks of sharing where there is little or no security or privacy protections in place.