Common bean is one of many legumes (family Fabaceae) widely cultivated for their edible seeds, seedpods, and leaves. Despite its benefits and life dependency especially in sub-Saharan Africa, due to climatic change, high population expansion, harsh weather conditions, and other disasters including diseases are causing considerable loss, where only foliar diseases are causing 20 to 80% of common bean production loss every year.

To overcome these challenges, bean-breeding programs were established prioritizing the development of new seed varieties to ensure high production and yield stability under adverse environmental conditions. High-yielding and resilient seed varieties are known from their observable traits which are measured through the process of phenotyping. However, development of improved common bean seeds is still a problem where it takes up to ten years to develop a new bean variety.

The major challenge Tanzanian bean breeders are facing is manual phenotyping, which is a tiring slow field process with manual paper writing as data collection tool and has many data errors, data loss, and data biases as results depend on the eyes of the viewer.

To increase and optimize the utilization of technology in agriculture for quick better seeds identification, this study introduces the use of artificial intelligence with hyperspectral data to help breeders in the classification of disease damage extent and timely plant health monitoring that will help them in the early identification of bean varieties resistant to foliar diseases and decision making using automated phenotypic data analysis reducing time used to spend on field and on manual data analysis.