ML ROBOTS
Professor: Soomeen Hamn
Guest Instructors : Damjan Jovanovic
Assistant Teacher: Laure Michelon
Class: SCI-Arc - Summer 2020- Design Lab
Team: Yibo Qiao, Isabel Socorro
Machines and architecture have had a relationship for a long time. Throughout history, every architectural style has been a reflection or a reaction of events, lifestyle, and technology of their time.
Now more than ever, technology is influencing the way we live, think, and in the case of architecture, the way we design.
The idea starts here
“M L Robots” presents a series of robots that can move around any architectural surface, in this case, architectural glass facades. The design lab research project aims to explore and investigate the development of an automated façade system which will provide a controlled environment for hydroponic systems.
Machine Learning Agents
MLAgent is a behavior-based artificial intelligence in Unity. The goal for this model is that the moveable agent can find the right plants and provide shading, watering and etc. There is a reward and punish system. When the agents find the right plant to cover, they get one point. Otherwise, the agents lose the points. Through the training, the agent would be able to find the right plants to cover.
The Sensing layer will get the environment data from the central computer. after receiving this information, the ML Robot will process this information to predict which plants need shading or irrigation. The center computer helps the agent to evolve their behavior and uploads the data to the user through the app.
