Researchers from MIT and Georgia Tech developed the first 3D model of mosquito flight based on visual and chemical cues, identifying three distinct patterns: a quick fly-by with visual cues alone, back-and-forth double-takes with chemical cues alone, and an orbiting path when both cues are present.

The study highlights public health implications, informing smarter trap design and contributing to data-driven, 3D mosquito behavior modeling.

An interactive app was released to let users visualize and adjust cues, observe simulated mosquito trajectories in real time, and explore trap configurations.

The work emphasizes calibrated multisensory traps to keep mosquitoes engaged longer, supporting broader efforts against diseases such as malaria, dengue, and West Nile virus.

Data were gathered from 20 experiments, yielding over 53 million data points and more than 477,000 flight paths, from which a simple yet accurate dynamical model emerged.

The three patterns are: fly-by when cues are visual-only, double-takes with chemical cues, and orbiting when both cues are present, with landing following the orbit.

In controlled experiments with Aedes aegypti, scientists tracked 3D trajectories using a black visual cue and carbon-dioxide as chemical cues, collecting more than 53 million data points and over 477,000 flight paths.

A key finding is that the combined multisensory response is not simply additive; the presence of both cues produces a unique orbiting trajectory rather than a simple sum of the single-cue paths.

The research aims to translate these movement insights into practical pest-control solutions by calibrating traps to engage mosquitoes longer and improve capture rates.

The model can be extended to include cues such as heat, humidity, and odors, enabling multisensory trap designs that improve capture rates.

Future inputs like heat and humidity could be incorporated to predict trajectories and optimize trap configurations for multisensory lures.

Funding came from NSF, Schmidt Sciences, the NDSEG Fellowship, and the MIT MathWorks Professorship Fund, with collaborators across MIT, Georgia Tech, UC Riverside, and related partnerships.