What investments are needed to reap open data benefits and what measures must be adopted to protect farmers from open data?

Open Data is already here, acknowledging it's potential and how to use such remains a challenge... Structuring data, raw data on initial investments for the last three decades, left us an aggregate database requiring, to date, soft and selective investments on research based/proven data over marketability, usability and alike.

Link http://comengip.org/en/ 

My actual interest and work remain in such areas of great interest to further widespread use of OD in Africa...

I am carefully watching the contributions in the forum and have to admit that there are some really interesting ideas presented. Just to add my two cents on the topic:

1. Farmers need to feel safe when openly sharing their data, and I do understand their unwillingness to do so currently. One of the most important investments that need to take place are national policies and legislation regarding the sharing of their data. Legislation needs to take into consideration existing work and aspects such as data ownership, data privacy, the sensitive nature of personal data, data sharing options, exploitation of data for commercial purposes, liability etc. If there is a thorough legal framework, than farmers will feel safer sharing their data. In this context, policy documents, guidelines, White Papers etc. on the management of (open) data, such as the ones developed by GODAN, CTA, ODI and other organizations, are valuable.
2. Investments in the data infrastructure. A country that wants to invest in open data and reap the benefits of it, will have to invest on a national infrastructure that will be responsible for data recording, aggregation (from various sources), management (organizations, sharing) and publication/sharing. Such an infrasstructure will also have to be open and interoperable, easy to connect to existing data tools (no need to re-invent the wheel) and easy to use. This would also need investment in terms of human resources, regarding people with the appropriate technical and open data background that will be able to maintain the operation of such infrastructure.
3. Training: Training in the concept of open data, their benefits and even the use of open data repositories, are essential for the successfull implementation of an open data plan. Data providers (e.g. farmers) need to understand the benefits and risks of sharing their data and how to share them. They need to be trained on the use of tools that they can use for recording data from their farms (both desktop and smartphone apps) and those data-powered tools that they can use for monitoring their farms or find the data they need from open sources. Investments in open data training are essential.
4. Long term plan: The exploitation of open data will take some time to show results, and it is important to have a long-term plan for this purpose. The plan needs to be flexible so that it can adapt to different conditions and challenges that may arise during the implementation of an open data plan.

There are more components in which investment will make sense, but I'll have to skip them as time is limited. I also agree with most of the points already made in this topic.

As rightly stated by Ajit Maru in his post in week 1 and 2, much of the available open data and information is not as useful as it could be, therefore not leveraging its full potential.

Innovative small startups have the potential to leverage that data. See the article about Frugal Innovation: the quiet revolution that is fighting off inequality. See https://www.weforum.org/agenda/2017/04/how-frugal-innovation-can-fight-off-inequality/

Many frugal innovations, like mPesa, are “empowering”.

This is also stated by the Open Knowledge Foundation.

“Open knowledge is what open data becomes when it’s useful, usable and used - not just that some data is open and can be freely used, but that it is useful – accessible, understandable, meaningful, and able to help someone solve a real problem.

So open knowledge is empowering – it helps us effect change and improve the world.” See

https://okfn.org/about/vision-and-values/

Governments and UN agencies need to make sure their data is more useful, usable and used, to facilitate open innovations. This is part of their core mandate.

In order to avoid that data and innovations are locked away, they need to be “unlocked” by using “Open Standards, Open Data, Open Source, and Open Innovation.” See http://digitalprinciples.org

The discussions about open data poses a dilemma. The difference between open data on the one hand and micro data on the other, collected through all sorts of digital sensors; e.g. mobile, social, IoT and satellite, much of it in real-time.

Where the former needs to be available in more useful, usable and used ways, the latter poses protection, privacy, ethics and ownership questions. Who owns the data “at the farm level”, now that digital sensors are collecting more and more micro data? Many studies reveal the potential of this big data and its capacity to steer innovation and provide new insights by using data science.

A recent report by GSMA and UN Global Pulse states that the “Achievement of the 17 Sustainable Development Goals (SDGs) will require development and humanitarian organizations to utilize new sources of data, technologies and skills for planning and programme implementation. The real-time digital data derived from mobile phone usage is among the richest data sources about population well-being. As this data is held by companies, the public sector cannot fully exploit it without leadership and partnership from the private sector. What we need is action that goes beyond corporate social responsibility. We need big data to be treated as a social good,” said Robert Kirkpatrick, Director, UN Global Pulse. See

http://unglobalpulse.org/news/state-mobile-data-social-good-report

Two balanced actions are needed; we need open data to become data as utility and big data to be treated as social good.

One practical way that farmer (or other data provider) rights can be protected without encumbering research with red tape barriers is through the use of open licenses (eg. the various Creative Commons licenses). These licenses essentially i) allow creators to assert copyright/ownership over their data while ii) permitting it to be freely used in certain ways.

For example, almost everything we publish goes out under a Creative Common by-attribution license. Anyone is free to redistribute, copy or build on our work, the only restriction is that they have to acknowledge the source. The software we write goes out under a GNU General Public License, which is quite similar but applies to code.

There are different types of Creative Commons license, and some restrictions can be applied such as "non-commercial" use or "share alike", depending on the wishes of the creator. If farmers wished to prevent application developers analysing their data and selling it back to them they could probably achieve that with such a license.

Open licenses are not without problems though. One of the main issues is that variations in conditions often make different "open" licenses incompatible with one another, so it becomes difficult or impossible to combine code or data that have been distributed under different licenses.