Real-time Simulation in Game and Reinforcement Learning
A participatory approach toward design build
Â
as a part of the
AIAAF 2023 Symposium -> LINK
3 days workshop in presence by CHIEN-HUA HUANG -> LINK
The workshop will be hosted at classrooms of the Udk Berlin
Clear your calendar - Huang is going to teach you Reinforcement Learning techniques in the Unity game engine sandbox. This workshop comes as a part of AIAAF 2023 and you will get basics in RL and ML-agents, virtual robotic and volumetric design. The workshop will be held from 3-5.May. and the number of the participants is limited to 15. The cost for the whole workshop is 120€.
Key Words: machine learning, coding, generative design, C#, python
Required Skills: Basic programming knowledge, 3d modelling, intermediate Rhinoceros, basic grasshopper, basic Unity 3D
Required Software: Unity 3D, Grasshopper; optional: Karamba, Anaconda
Workshop Type: Tutorial
This workshop delves into the intersection of real-time simulation in Unity game engine and AI as a unique method for architectural participatory design and construction. With the rapid advancements in machine learning and AI in the architectural industry, there is a need for a more accurate rationale and a wider spectrum of evaluation in the design process. By utilizing machine learning and Unity3D for game design elements, there is potential for promoting active participation from architects, artists, and the public in the design generation process by merging gamification and design evaluation. Reinforcement learning, a subcategory of ML, is also being increasingly explored in the design industry due to its interactive nature. In this workshop, we will use state-of-the-art techniques in Unity to generate methodologies for volumetric design and simulation of virtual machines via working with multiple platforms. Through this workshop, we will examine questions such as how humans and AI can collaborate to design creatively and inclusively, how AI-made elements can be reiterated and fabricated through robotic simulation, and how data such as structural evaluation and environmental performance can be streamed and interacted with in real-time to involve stakeholders and AI. These questions align with the goals of AI proponents for the new generation of data processing to aid designers through the novel augmentation of machine collaboration and simulated digital fabrication.
