Gesture as Prompt
Project Type | Individual Research |
Year | 2024 |
Location | London |
The project tryies to extend beyond the conventional applications of modeling in architecture, aiming to redefine the model's function from being merely a design or visualization tool to a multifaceted instrument capable of inciting systemic change across various industries. By emphasizing that 3D geometrical changes possess vast possibilities and can be significantly more dynamic than traditionally envisioned, It tries to draw an image of revolutionizing our approach to understanding and interacting with any system through 3D visualization and advocating for a paradigm shift in the perception and utilization of models, proposing their use as dynamic tools for exploration, analysis, and systemic innovation beyond the architectural. domain.
​
Traditionally confined to the realms of design and visualization within architecture, models hold untapped potential for broader applications, suggesting that stakeholders from diverse sectors could harness this potential to address complex challenges.
OMA's Foam Models for Testing Idea
Building Information Modelling (BIM)
A technology enables the integration, visualization, and management of information across the entire building, viewed as a complex system
If the idea of using model as a tool to synthesize information in a system (i.e. BIM) is applied to other system, how would it look like?
This initiative explores the possibility of models serving as interactive platforms for AI to communicate multidimensional changes within systems like supply chains, thereby pushing the boundaries of what models can represent and accomplish. It pioneers an innovative approach to reimagining supply chains through the lens of three-dimensional visualization, utilizing space as a transformative tool for systemic change.
cost per unit of
economic gain
processing time
transportation cost
By converting the traditionally two-dimensional representations of supply chains into a dynamic 3D model, this project allows for the integration of multifaceted parameters, offering a holistic view of the system. For instance, transforming the UK's beef supply chain into a 3D model with cubes representing various segments, where the y-axis signifies transportation distance, while the x-axis and z-axis illustrate key supply chain parameters such as the process time in this sector and cost per unit of economic gain, respectively.
3D Beef Supply Chain Demostration Video
This 3D framework not only clarifies the physical dimensions of supply chains (y-axis) but also encapsulates the flow of time (x-axis) and cost structures (z-axis), thereby providing a comprehensive perspective for analysis and optimization. Movements along the y-axis reveal shifts in transportation distances, extensions along the x-axis indicate variations in time requirements, and expansions along the z-axis reflect changes in costs. This multidimensional approach not only aids in understanding how individual variables impact the entire system but also identifies the interrelations and interactions among multiple variables.
Leveraging a large language model trained with extensive supply chain information, the project has developed an AI intimately familiar with the internal relationships within supply chain systems. It plays a crucial role in presenting changes within the supply chain blocks. When modifications are made to a specific block, such as moving the consumer segment closer to the farm segment on the y-axis, the AI predicts and communicates the resultant 3d alterations of other blocks, including their potential expansion, reduction, or even disappearance, along with their respective changes on the x, y, and z axes.
Test Motion and Scaling in AR
We can imagine this model could be constantly in motion by being given simple gesture to any block of it. Then diverse geometric changes beyond simple scaling, such as various complex geometric transformations may lead to more unexpected systemetic results. And we see the possibility that gestures could serve as prompts. This inquiry delves into the existence of a three-dimensional prompt that transcends traditional modes of communication like text, language, sound, and imagery, leveraging the language model's capacity for concrete interpretation of abstract 3D gestures, much like how image AI can interpret random abstract pictures with its own understanding.
How AI understands the gesture of twisting, hollowing and scaling 1D in the farmer layer of the supply chian?
Various Gestures in Architectural Modelling
GALLERY
(CLICK TO VIEW IN HIGH-RES MODE)