Proximal and Physical Sampling with Agricultural Mobile Robotics

Abstract:
Agricultural mobile robotics play an increasingly significant role in precision agriculture.
Besides remote sensing with aerial and ground robots, proximal sensing and physical interaction
with the crop are gaining momentum as they can offer localized data to enable more informed
agriculture support systems. Yet, proximal and physical sampling with agricultural mobile robots
demonstrate challenges and opportunities. One relates to determining where to sample. Often
times, prior information maps (e.g., field soil moisture maps) are available but due to the
dynamic nature of the problem these maps may not be trustworthy. This calls for task and motion
planning methods that can harness prior information but not over-rely to it. The other relates to
determining how to sample. It is not always the case that sensors for proximal sensing or
(custom) actuators for sampling can fit to an existing robotic platform. This calls for an
integrated design of actuation and perception to co-optimize robot performance/mobility and
sampling.

This talk will cover some of our recent advances helping address determination of both where
and how to sample. We will discuss a new task planning under uncertainty algorithm for
precision agriculture applications whereby task costs are uncertain and the gain of completing a
task is proportional to resource consumption (such as water consumption in precision irrigation).
The goal is to complete all tasks while prioritizing those that are more urgent, and subject to

diverse budget thresholds and stochastic costs for tasks. Furthermore, we will discuss new co-
designed means for proximal sensing and physical sampling. In proximal sensing, we will

discuss a robotic means to perform field apparent soil conductivity measurements and compare
them to maps created via use of current standards. In physical sampling, we will discuss about
the design of a mechanism for insect capture/release that it can be mounted on aerial robots as
well as of a leaf retrieval mechanism to identify and cut leaves at their stem in tree crops.

Bio:
Dr. Karydis is an Assistant Professor in the Department of Electrical and Computer Engineering

at the University of California, Riverside (UCR). Before joining UCR, he worked as a Post-
Doctoral Researcher in Robotics in GRASP Lab, which is part of the Department of Mechanical

Engineering and Applied Mechanics at the University of Pennsylvania (Penn). His work was
supported by Dr. Vijay Kumar, Professor and Nemirovsky Family Dean of Penn Engineering. He

completed his doctoral studies in the Department of Mechanical Engineering at the University of
Delaware, under the guidance of Dr. Herbert Tanner and Dr. Ioannis Poulakakis.

Dr. Karydis’s research program addresses foundational robotics research problems underlying
applications in emergency response, environmental sensing, precision agriculture, and (more
recently) pediatric rehabilitation. His research seeks to enable existing and new robot
embodiments (and teams thereof) to operate in efficient and resilient manners autonomously
and/or in cooperation with humans despite the presence of uncertainty associated with action,
perception, and the operating environment.