S18 - Session P4 - Can fruit walls raise the prospects of sensing and robotic technology in apple production?

Authors: Lars Zimmermann *, Cory Whitney, Eike Luedeling, Martin Balmer

Apple growers are increasingly interested in applying sensing and robotic technology to enhance the efficiency of their production processes. Such technological assistance may hold potential to reduce pesticide applications, lower operating costs and increase labor efficiency, among other favorable outcomes. Since complex tree architectures pose technical challenges for both sensors and robots, such developments are often accompanied by calls to simplify tree structures to make them compatible with emerging technologies. The objective of this study is to explore the potentials and challenges of narrow fruit walls for optimizing the effectiveness of sensors and to analyze the economic efficiency of such cultivation systems. We evaluated the influence of tree shape on the detection of phenological characteristics using a sensor platform consisting of three RGB-D cameras and a LIDAR sensor. Since one growing season, we are applying this setup to detect blossoms and fruits in seven different growing systems. Our aim is to evaluate whether tree shape has an effect on the ability of sensors to detect relevant features within the canopy. In further analyses, we are planning to evaluate the viability of the tested training systems from a plant physiological and economic point of view. Growing systems will be evaluated using decision analysis approaches, which can represent all relevant influencing variables in a multifactorial model. Decision analysis can scope out the economic feasibility of innovations by projecting the full range of plausible outcomes through probabilistic simulations that fully consider all relevant uncertainties and account for all pertinent knowledge gaps. Applying these methods will enable us to generate actionable advice to apple growers on whether a transition to fruit wall training systems is likely to enable more efficient production through sensing and robotic technology.