Development of the Fynbos Leaf Optical Recognition Application (FLORA): An innovation journey of a tool to assist in identifying plants

Abstract

The Fynbos Leaf Optical Recognition Application (FLORA) is a software program to automatically identify fynbos plants using leaf photographs. While it is easier to classify fynbos when they are flowering, most fynbos flower for only short periods therefore FLORA was designed to identify plants by leaves instead of flowers. This paper presents the innovation journey of FLORA, highlighting transitions in development spaces, impact of requirements changes, and other significant challenges and lessons learned in the journey. The development was done out in a university research context and vacillated between being in a closed space and being a more open initiative. The project settled on being a collaborative and open innovation whereby the system supports a more diverse community of users and contributors. While the original requirements concerned a small scientific community of students and scientists botanists, the revised system, which the innovation journey lead towards, aims instead towards a wider community including tourists and schools pupils. It is hoped the innovation will have a broader societal influence in particular at schools level, where it is hoped that FLORA will both inspire young learns, and in particular tech savvy kids who spend too much time indoors, to spend time outdoors and to improve their awareness and appreciation of nature. This paper concludes with ways the project could have been streamlined from early on to better support the users and to facilitate the transition from a close to an open innovation.