Vinobot and Vinoculer 3D model created by Vinobot 4D-RGB model created by Vinoculer 3D model created by Vinoculer

Ali Shafiekhani


I am a Ph.D. candidate at the Vision-Guided and Intelligent Robotics Laboratory (ViGIR-Lab) in the Department of Electrical Engineering and Computer Science at University of Missouri-Columbia under supervision of Dr. Gui DeSouza.

Since I joined the ViGIR-lab in late 2014, I have been working on developing robotics platforms for plant phenotyping in agricultural fields.The skills that I acquired during the first years of my PhD in the EECS department together with my background in Mechanical Engineering (Bachelor’s and Master’s) allowed me to develop a new architecture for field phenotyping that has been cited several times in different scientific and news magazines (e.g. WIRED , one of the largest, if not largest, printed and online magazine for “techies”, TechCrunch, EurekaAlert, Nature International Journal of Science , and many other plant-science related media, see NSF EPSCOR newsletter).

The architecture consists of two robotics platforms, an observation tower and an unmanned ground vehicle (UGV) to monitor plants at two scales, the entire field and individual plants. The architecture can extract important features and create 2D, 3D and 4D models that include spatial, time and heat signatures, all made possible by processing stereo RGB, Thermal cameras and multiple environmental sensors.

I also had the chance to teach senior and graduate students in two different robotics courses as the lab instructor. During those experiences, I integrated the Robot Operating System (ROS) into the lab assignments and helped students to learn new tools and understand basic concepts in robotic vision.

Currently I am working on the autonomous navigation of UGV in agricultural fields using computer vision. The challenges derive from the disturbances and uncertainties (e.g. wind, dynamic scene, illumination) and the requirement for high accuracy (i.e. 1-2 centimeters) to navigate without damaging the plants. The proposed approach in my research involves a distributed, multi-agent algorithm for localization of robots in the field.

In the past, I worked as a research assistant at the Center for Mechatronics and Intelligent Machines (CMIM) in the Mechanical Engineering Department at University of Tehran where I designed and implemented an adaptive neuro-fuzzy controller on an unmanned bicycle.