Anisha Nakagawa

MIT Department: Media Arts and Sciences

Undergraduate Institution: Olin College of Engineering

Faculty Mentor: Prof. Danielle Wood

Research Supervisor: Keith Stober

Websites: Personal Website, LinkedIn

2018 Research Poster


I’m a student at Olin College, originally from Cambridge, MA. I’m interested in using electrical engineering, computing, and human-centered design to create sustainable, resilient systems and empower people. I also enjoy cooking, biking, and exploring new places.

2018 Research Abstract


Sensor networks and complex system modeling to monitor and predict water hyacinth growth on Lake Nokoué, Benin

Anisha Nakagawa1, Javier Stober2 and Danielle Wood2

1Olin College

2Space Enabled Research Group, Massachusetts Institute of Technology

Water hyacinth is a widespread invasive aquatic plant, and it is problematic in the region of Lake Nokoué in Benin where it blocks waterways, hinders fishing and travel, and creates an environment for breeding malaria-carrying mosquitos. Green Keeper Africa, a local company, has developed a method to turn water hyacinth into a material that absorbs oil-based waste. The company developed a business model around hiring local people to collect water hyacinth, and they are interested in using engineering methods to detect and predict where water hyacinth will grow. In support of the vision of Green Keeper Africa, this research investigates two methods for monitoring water hyacinth. Low-cost water quality sensors are used to collect salinity, temperature, turbidity, and dissolved oxygen information, which are then processed with a microcontroller and transmitted wirelessly to analysts. Agent-based modelling is used to simulate the interactions between human behavior and water hyacinth growth, which provides a platform to test different hypotheses and situations. These techniques are found to provide more insight into water hyacinth growth in Benin, and the methods could be applied to monitoring other types of aquatic vegetation in other geographic contexts.