.svg)
How might we design digital phenotyping tools that truly meet the needs and workflows of bean breeders in Africa?
User Personas and Experience Maps to guide the design of a mobile phenotyping App
In a world where the severe impact of climate change and population growth strain global food systems, the ability to rapidly breed resilient, high-yield crops is more critical than ever. Phenotyping - the science of observing plants and selecting standout traits for mulitiplication - is a cornerstone of crop breeding. Image recognition and Artifical Intelligence (AI) unlock new possibilities to support this process, capturing plant data at unprecedented scale and precision. Yet, despite this leap forward, the technology remains out of reach for local breeders, held back by high costs and limited access. The Artemis project aims to enhance bean breeding programs in Africa with AI-powered phenotyping. To ensure the technology effectively supported breeders, the teams' initial focus was deeply understanding the users and their context in Tanzania. We supported this process with exploratory fieldwork and visualizing insights to guide the development team.
We conducted one-on-one interviews with breeders and their teams, complemented by direct field observations during their daily work processes. Synthesizing these rich insights, we developed three distinct User Personas - archetypes that provide insights about how groups of users behave and think, what they want to accomplish, as well as their motivations. Another key step in understanding breeding practices was mapping the breeding workflow for each User Persona. This helped us gain a clearer picture of how breeding programs operate, including the stages involved and the decisions made along the way. By visualizing the full process, we were able to unpack the complexities of phenotyping and ensure that all critical elements were considered.
1. Who are the primary users of potential digital phenotyping tools?
2. What are their day-to-day workflows, challenges, and motivations?
3. How can digital tools be effectively integrated into existing breeding practices?
Breeders face significant constraints in using digital tools for their phenotyping, as for example resource scarcity, lack of digital infrastructure, or digital literacy. Different roles (lead, coordinator, technician) have distinct needs, perspectives, and concerns regarding new technology (e.g. regarding efficiency gains vs. learning curve, or data privacy). Understanding the full breeding cycle context is vital for designing relevant tools. Typically, a breeding program spans around 10 years and includes four main phases: (1) market segmentation, (2) crossing and screening, (3) breeding, and (4) variety release—the final outcome of the phenotyping process.
The User Personas and Experience Map provided a shared understanding of the users, fostering effective communication within the Artemis project. This human-centered foundation ensured that the subsequent development of the phenotyping technology was directly tailored to breeders' preferences and goals, increasing the likelihood of successful adoption and impact on breeding efficiency and accuracy.
Our work supported the creation of three detailed breeder Personas as well as the development of a comprehensive breeding workflow Experience Map.
.jpg)
