BHT Berlin & Umweltbildungszentrum Listhof e.V.

KInsecta · Citizen Science

How an AI learns to recognise insects
The product

KInsecta records insects with a multi-sensor system: optical, mechanical and acoustic signals instead of a net. In a web application, the project team reviews the measurements and identifies the species. The result is cleanly labelled datasets – the basis on which automatic insect recognition is trained. Behind the research project stand the Berliner Hochschule für Technik and the Umweltbildungszentrum Listhof.

  • Research project
  • Citizen science
  • Insect monitoring
  • Open source
  • Web application
  • Multi-sensor system
Our part

One application, three stages

We gathered the requirements together with the project team and built the web application – from the prototype through two stages of expansion to the public map view. Added to that: test and live servers, the development toolchain with automated deployment, and setting up the monitoring integration. Until the end of the project funding we kept servers and frameworks up to date. Today the community keeps collecting its insect data in this application.

Services
  • Requirements gathering
  • New development
  • Full-stack development
  • Development toolchain
  • Public API
  • Servers & deployment
Tech stack
  • Java
  • Spring Boot
  • Vue.js
  • Single-page application
  • MariaDB
  • S3 storage
  • Docker
  • GitLab CI/CD
  • Ansible
  • SonarQube
  • Datadog
Got a product in mind? Let’s talk.