Overview
The NREL project team will develop a software system for specifying prefabricated building retrofit components and apply the system and corresponding prefab components to a low-rise multifamily development. The design will be tailored to each building using a machine learning model applied to site scan data, and the fabrication process of insulation panels and HVAC components will be done offsite with support from 3D mixed-reality tools.
| Project Type | Problem to Solve | Solution | Location | Timeline | Partners |
| Residential retrofit | Time, cost, and intensive detailing barriers of deep energy retrofits | 3D mixed-reality tool for streamlined design, fabrication, and deployment of retrofits | Salt Lake City, UT | October 2022 – March 2027 |
|
Project Goals
The goal of this project is to develop a scalable system that reduces the cost and installation time and increases the energy efficiency and durability of deep energy retrofits for low-rise residential buildings. This project specifically targets the substantial labor effort required at each stage of the retrofit project life cycle to customize each building subsystem for each building to be retrofitted.
Impacts
Automation and offsite prefabrication of deep energy retrofit hardware can yield substantial retrofit cost reductions and quality improvements. This project will demonstrate these benefits in the United States, where residential buildings vary widely in form and other characteristics, making simple, rapid, building-specific customization capability critical to scaling the adoption of deep energy retrofits.
Technology Impact
The workflow includes a remote expert for real-time installation guidance and is expected to reduce construction time and project costs by 50%.
Market Impact
The outcome of this project will be reduced material waste, reduced project costs, reduced installation time, reduced occupant disruption, and superior quality, resulting in improved market acceptability, scalability, and uptake of deep energy retrofits.
- 2023: Site and scope were identified.
- January 2024: Eight occupied units were chosen for monitoring.
- February 2024: Building scans were taken to be used in the machine learning model and for building component manufacturing.
- February 2024: Equipment was installed in the eight selected units to monitor indoor air quality, comfort, and energy use pre- and post-retrofit.
- Monitoring occupied homes necessitates a less visually intrusive and more robust instrumentation suite as compared to monitoring unoccupied homes or test huts.
- Recent rain or snow, in addition to cold or hot outside temperatures, is very useful for thermal-imaging diagnostics.
- Training machine learning systems for façade and ceiling object identification requires a well thought-out hierarchical classification structure.
- NREL will be monitoring pre-retrofit data and creating energy models to inform the wall-panel design.
- Trimble and NREL are collaborating to develop machine-learning algorithms to be used in translating building scans to panel production.
- FunForm and IBACOS are coordinating product and material specifications, as well as factory assembly.
- Energy and hygrothermal models will inform optimal panel design, both at the demonstration site and at other locations throughout the United States.
General Inquiries: [email protected]
About the ABC Initiative
The Advanced Building Construction (ABC) Initiative, led by the Building Technologies Office (BTO), integrates energy efficiency and advanced technology solutions into industrialized construction processes to drastically increase the speed and scale of high-performance, low-carbon building retrofits and new construction.
.....
Page last updated: June 5, 2024