unit-code
Under the background of the pandemic normalisation and the second digital turn, Tesseract is committed to meeting all the requirements of the occupant in a tiny scope of life through big data computation and voxel-based material system.
Tesseract is a real-time adaptive customized living architecture system, which gets rid of the inherent life cycle and rigid form function through the flow of data, continuously reshaping intelligent multi-dimensional communities.
In the era of big data, Tesseract sets up a comprehensive information collection port to monitor the status of users and the environment and establish a complete database. In the face of the diverse requirements in the big data era, a novel agent-based spatial planner algorithm was developed using reinforcement learning to provide adaptive policies for adjusting volumetric room boundaries to constantly changing scales, shapes, materials, and atmospheres. Real-time negotiation balances the interests of multiple users through a socio-economic model and puts them in a dynamic equilibrium state. In addition, relying on the autonomous material system, Tesseract completes the independent construction and reconfiguration of space and realises the real-time linkage between the occupant and space.
Welcome to the world of Tesseract!
Enjoy adaptive customised life and autonomous smart community in Tesseract.
At the stage of pandemic normalisation, people are forced to limit themselves to a tiny scope of life. Have you ever been tired of your room and missed the big desk in the office? Tesseract can help you.
The artificial sensing system and interaction platform collect the environment and user information, transmitting it to the console to guide the real-time reconfiguration of the material system.
Tesseract has a cyclic, adaptive, and autonomous workflow.
In the material system, the voxelised static components provide the minimum resolution, making it an ultimate discrete system with maximum variability.
Components of different materials and forms make up a library to meet various scenarios. The replaceable panels can also be customised with traditional materials to blend into the local built environment.
The annulata robot has an efficient sliding motion mode, which does not require additional operating space, and in most cases only needs to overcome rolling friction.
L Mode is the most efficient combination pattern, It has many variations in the coordinate system.
Raspberry Pi and U2D2 aim to collect and process data. Python and C# are used to create a set of codes for controlling the movement of the motor so that the physical prototype can get a digital twin in unity.
The space generation algorithm can compute specific solutions in real-time, maximising space utilisation. It has both bottom-up local intelligence and top-down global constraints.
The space schema is the quantification of the space occupation behaviour, mainly consists of three parameters: volume, proportion, and form.
The negotiation schema reflects the features of the agent's interaction with the external environment and other agents.
The base map represents the environmental influence on the agents. The masking map is the global constraint of the planning scope by the structural support.
An experiment in base station No.226 shows how the space generation algorithm works in different situations and occupied rates.
The construction sequence shows how the material system realises the algorithm space. The two kinds of components continue to lay tracks for each other, sliding relatively to adapt to real-time changes.
Reinforcement learning is used in the construction algorithm to make the robot autonomous.
The robot can reconfigure the static components in 3D based on interactive instructions. Users can ask the robot to build small furniture or open doors and windows.
The mobile platform system directly connects with users and undertakes the functions of information collection, storage, analysis and recommendation.