Skip to content

Yannick947/deep_bau

Repository files navigation

DeepBau

Hackathon for construction sites 48hours, 2 persons

Video pitch and full description available on Devpost.

Concept

"DeepBau concept"

Inspiration

Digitalization is not bypassing the construction industry. Everywhere you hear buzzwords like BIM or lean construction. These methods are used to optimize the individual process chains of construction projects and to identify interfaces as early as the palletization phase. But where exactly do problems arise? How can we avoid wasting resources such as time, material and manpower? This is exactly where we want to start with DeepBau. There is a lot of data, but it is stored unused on servers of individual companies. We evaluate this data and try to develop optimized construction process plans with the help of AI. Our process mining tool DeepBau brings structure and clarity to confusing construction projects. Lean construction is made possible through smart visualizations of process chains and AI algorithms are used to detect future events and problems at an early stage.

"DeepBau concept"

What it does

User-defined visualization for process chains in ongoing construction projects Visualizations of the latest progress on the construction site Visualization of the temporal course of resources and processes: Work equipment, material, people, weather AI algorithm to predict future project steps

How we built it

  • Django webapp
  • Bootstrap v5 frontend
  • Python backend
  • LSTM encoder-decoder architecture

What's next for DeepBau

Direct connection of the api of the datalake to our webapp (currently all existing data is downloaded and loaded from memory) Improving the training of the AI algorithms with more computing power.