Introduction to BenchBot

High-throughput technology for plant phenotyping allows plant development and morphological traits tracking at a scale that was unimaginable with traditional phenotyping techniques. The development of high-throughput phenotyping tools has been exponentially growing for several years now thanks to the increasing availability of lower cost sensors and technological advances in general. There are numerous examples of these tools which are being used for different applications such as root, shoot, and tissue phenotyping. A review article by Yang et al. (2020) presents an excellent summary of the evolution of these tools as well as applications and limitations (link below). This article shows how even though there are several options out there, most of them are either costly, need specific infrastructure, are limited to certain plant sizes or require large amounts of highly trained labor to operate. This is the gap that we are expecting to cover by developing a a versatile low-cost high-throughput tool such as BenchBot.

BenchBot (BB) is an open-source high-throughput plant phenotyping computer vision solution which allows for data collection in an automated way, it can be used in different setups (indoor/semi-outdoor) and with various sensors. It is relatively low-cost, modular, upgradeable, portable and easy to use.

In this space you will find all the information about this project organized from most general to more specific. The sidebar on the left will help you to easily navigate from one section to the other. You can read the entire document or skip to the sections that are of most interest to you.

Within this space you will find links to google spreadsheets and docs as well as to GitHub repositories which complement and complete the documentation that you find here.

Semi-Field setup for weeds image collection at the Beltsville Agricultural Research Center (BARC), ARS, USDA in Maryland