Quick and “dirty” insights from the internal CIAT conference on ICT4D

The path to progress for ICT4D at CIAT: some insights to help shape the way forward Before CIAT’s KM group engages in a deeper documentation and reflection on the experiences that were shared during CIAT’s internal conference on ICT4D, I decided to write this preliminary summary of my takeaways.

The internal call for proposals resulted in the submission of 15 abstracts which were combined into 4 thematic sessions. Many thanks to Sophie Treinen, Knowledge Outreach Team leader at FAO, who gave a virtual key-note and summarized the lessons learnt of 10 years of E-agriculture- the findings of which were recently published in FAO e-agriculture report.

Find all presentations on slideshare: https://www.slideshare.net/CIAT/tag/ciat-ict4d

10 years of E-agriculture: aligning technology with end users’ preferences The report’s recommendations highlight 7 key areas (Content; Capacity development; Gender and diversity; Access and participation; Partnerships; Technologies; Economic, social and environmental sustainability). These are hardly a surprise, but what struck me was that the presentations made during the two day event addressed many of these lessons learnt, dos and don’ts. ICTs and communications for development: the struggle with scaling outwards and upwards Participatory video related to farmers’ perceptions of climate change in Central America. This project was moreso conceived as a research project on ICT4D. One of the main challenges of the process is scale. Participatory video is time-consuming and limited to a small number of participants. Thus, as Manon Koningstein so aptly pointed out, how can we reach other communities and farmers with a product that is, limited in its quality because it’s produced by amateurs? Read the working paper of the CGIAR research program on climate change CCAFS.Farmer extension on cassava pest and disease management in Asia. Kris Wyckhuys- CIAT’s cassava entomologist had previously used videos for farmer to farmer extension, and thus believes in this media. While we do know that the video was well received, a concrete evaluation about its usefulness is not yet available. Louis Parker and Lan Le talked about bottlenecks; one of which I recall was on language issues while others included the video being distributed and projected in different regions of different countries, and changes in accents and dialects, which are issues that not only affect the understanding but also the credibility of the information being disseminated. Read more on this blog post from the CGIAR research program on Water, Land, and Ecosystems. ICTs for data and information sharing among local stakeholders: easier said than done!

Tweets from conference

The four presentations in this session confirmed that stakeholder engagement in the realm of ICTs is much more difficult than expected.  The culture of consistent data collection and analysis is far from being wide spread in rural communities, where decisions are often made based on tacit knowledge. These are the results that Fanny Howland presented, and it’s not good news for Sylvain Delerce and the team who work on site-specific agriculture. The two-way information system that he showed us works best with industrial farmers and even then, the amount, quality and consistency of the data sets sometimes limit the construction of a robust decision support tool. The team si now looking into remote sensing techniques to solve the lack of consistent data capture by smallholder farmers. Anton Eitzinger and his team who work on climate-smart agriculture follow a concept of collaborative problem solving by connecting “outside-expert” and “local-expert” knowledge, through the use of an android application which interacts with a web-based platform. Knowledge brokers then provide the human interface. Read more here. At the almost opposite end of the spectrum, lies the online information system that Shadi Azadegan and Alejandra Mora presented; as it aims at keeping stakeholders engaged in the creation and use of project –related information. A comprehensive system was designed and developed from scratch, and still needs further development- both in terms of its applications and user training.  This then begs the question: does a home-made platform such as this have a chance of survival once the projects and programs come to an end? Data collection using mobile devices: should we aim for the power of the collective? Another four more presentation took us through this thematic session. Angela Hernandez guided us through a solid design and implementation process of a home-made mobile computing system for CIAT’s genebank. I retain her message about the essential step of prototyping and involvement of the end users- the field workers who had to overcome initial worries and fears. Daniel Debouk is CIAT’s genebank manager. According to him, the tool  presented “goes in line with our dream: farmers can order germplasm from a cell phone directly from their plots wherever they are on the planet, after consulting our web site with germplasm data, also from a cell phone.” The two examples of bean projects in Africa illustrated ready-to-use solutions such as the Open Data Kit, which generated an interesting and needed discussion on what solutions to select, adopt and promote among (even CGIAR) staff for easy use. Meanwhile, the 5Q approach is a really fascinating example of how a simple idea expressed by DAPA’s Andy Jarvis became a winning solution. This approach helps to engage in quick feedback loops with next and end users that then allow for the making of on-the-go adjustments during the project’s execution.  The difficulty however, lies in maintaining its simplistic nature. ICTs for data management and sharing with scientists and partners: a plethora of possibilities! We heard a captivating presentation by Daniel Cruz on a 12-year old effort to build a laboratory information management system (LIMS). The pitfalls, since the inception of the initial idea have been numerous, and the lessons learnt hard to swallow. While many obstacles have been overcome, the system is still not heavily used and its future quite uncertain. The Pan-African Beans Alliance demonstrated the use of their database which covers information from more than 7 areas of the network’s intervention including seed systems, capacity building, and nutrition. Innocent Obilil took us through this significant effort for a network that encompasses 29 national agricultural research institutions, and over 350 partners from farmers and rural communities, NGOs, traders and the private sector. The SOFT (Selection of Forages for the Tropics) web-based application is a pioneer effort and result of collecting tacit knowledge from forage specialists around the world.  While it is still heavily used with more than 180,000 visits a year, it needs to be updated, to not only survive changes in technology but also so that it can be available in more formats. The cassava genomics team who closed out the session, showcased two very different efforts on data management and sharing. On the one hand, the cassava genomics hub -an online tool for geneticists, and on the other hand an internal cassava data management site where the team gathers all historic and current cassava data for future publishing. Sustainability: how can we make ICT4D solutions last, or are they not supposed to? Our Decision and Policy Support research area might be well placed to lead the former. The data, information and knowledge group certainly hopes to provide assistance with the choice and implementation of ICT4D solutions. One burning question among the many posed however, is that of sustainability, together with some related questions which emerged during the discussions, such as:An emerging insight in the wake of this two-day event is that we need to distinguish the research we may wish to undertake on ICTs4D, from the use of ICTs as enablers of our research projects.

  • How do we save time in choosing and implementing ICT4Ds?
  • How do we make our ICT4D investments last?
  • Make vs. buy: what ICT4D choices are best in which conditions?
  • How do we reach the biggest number of users?

A more detailed documentation will be available soon.