Welcome to my research page. I am a PhD candidate and researcher at Ghent University in the Department of Electronics and Information Systems (ELIS). I am also a member of IDLab, imec and ILVO. My research is about non-linear dynamics of plants for computation and sensing.

Some non-technical questions we hope to answer by means of this research:

  • What plant is more intelligent: strawberry, orchid or corn?
  • What kind of environment does a plant like?
  • Does leaf movement tell anything about the plant's feelings?

Since my research has roots in both plant science and computer science, I work together with a number of different people from both ELIS, ILVO and BIOMATH.


There are already over 7.5 billion people on Earth. The world population is still increasing, and will probably continue to do so for several decades. This ever growing population in combination with global warming, results in enormous challenges in agricultural production around the globe. Agriculture will need to increase its production capacity to meet ever changing demands (both in developing and developed economies) while decreasing its environmental impact.

Precision agriculture is seen as an answer to address both issues of increased productivity and decreased environmental impact. Precision agriculture refers to a set of techniques that attempts to apply production factors in the right quantity at the right time. These factors include water, fertiliser and protection agents.

However, current techniques rely mostly on indirect measurements such as temperature, relative humidity, soil moisture level and nitrogen level measurements. They only indirectly indicate the state of the plant. As such, they are unable to accurately enable us to measure the exact plant state. We are going to directly observe the plant using a wide variety of imaging and contact sensors.

Research Goals

The research I am performing aims to unify the fields of physical reservoir computing and plant science. While plants lack a central brain-like organ, they do respond to an ever-changing environment. It has also been shown that plants exhibit memory, further indicating that plants possess embodied intelligence.

As such, we can identify the plant as a reservoir that implements a complex non-linear dynamical function. A reservoir is a randomly interconnected set of nodes (a recurrent neural network). This network is fixed and a linear readout function is used to map the network state to the desired output variables. The inputs of the system are the biotic (e.g. plant-to-plant interaction and pests) and abiotic (e.g. temperature, illumination and water availability) environmental factors. A schematic overview is provided in the figure below. In a first experiment, we want to validate this hypothesis.

plant as reservoir
The plant as a non-linear dynamical system. From left to right: the input, reservoir and output layers. The input layer contains all variables that influence the plant (e.g. temperature, water, pests, sunlight). The reservoir layer is here the plant. We will model this using a random (fixed) recurrent neural network. This layer implements the non-linear dynamic function. The output layer implements a linear readout of the observed data (from e.g. cameras and contact sensors). This readout then results into the targeted output values.


As of September 2017, I am involved in the bachelor course of Digitale electronica. This is an introductory course on digital electronics taught to Electrical Engineering and Computer Science students. I mentor the lab sessions during which the students need to programme an FPGA in VHDL.

Additionally, I am also involved in the course of Embedded Prototyping, a master course for the Electronics and ICT students in Kortrijk. In this course, students need to develop a design based on a request from industry.