Assessing building-level retrofit potential instantaneously and on a large scale
Research Team
Our Motivation:
"My experiences in the residential solar sector have showed me that effective building decarbonization efforts need to empower property owners to make sensible investment decisions. By identifying building-level retrofit potentials on a large scale, this research aims to facilitate the matchmaking between property owners and contractors/investors in the retrofit industry to help owners understand which retrofit investments should be considered."
Research Contribution
Provide a new methodology to bridge the gap between bottom-up and top-down building retrofit approaches.
Problem
Practical Problem
We are missing a method which can estimate building-level retrofit options fast, accurately, and on a large scale.
As a result, property owners are often unaware of their respective retrofit potential and contractors struggle to identify the most promising retrofit targets.
Solution
Explore the development of a scalable and data-driven solution to assess the retrofit potential of a building’s roof, walls, and windows solely based on publicly available data sources.
Added Value For The Industry
Demonstrate how publicly available data sources can be integrated to assess detailed building retrofit potential on a large scale.
Empower investors/contractors to approach owners proactively.
Cooperation Partner
Timeline
| Date | Activity | Outcome |
Summer 2022 | Research became awarded: Assessing building-level retrofit potential instantaneously and on a large scale | |
Fall 2022 | Literature review | |
| Data collection | |
| Data cleaning | |
| Model design | |
Winter 2023 | Implement baseline and deep models | |
| Conduct experiments | |
Spring + Summer 2023 |
|
Project Summary
(Provides you with a brief and clear summary of the insights and outcomes at the end of the funded year.)
Current methods to assess the energy efficiency of buildings require on-site visits of qualified experts or computationally expensive physical energy models, at times even a combination of both. This hampers our ability to analyze, monitor, and prioritize existing buildings in terms of their renovation potential as existing methods are slow, costly, and geographically incomplete.
To improve the energy efficiency of the building population with targeted and strategic renovations, we propose to classify the energy efficiency of individual buildings from widely available remotely sensed data sources such as street view, aerial, and satellite-borne thermal infrared imagery.
In an initial case study in England, we confirmed the feasibility of the proposed approach, outperforming existing heuristics for the remote evaluation of properties in terms of energy efficiency by 10-38%. In a follow-up study in Denmark, we proved the generalizability of the proposed approach by showing that the amount of information in street view and aerial imagery is consistent across countries.
Motivation to continue the project:
- Understand the role and value of additional data sources, such as LiDAR, which provide information on the 3D shape and geometry of buildings.
- Understand the role and value of data sources at different spatial resolutions, e.g., by using high-resolution thermal infrared data.
- Improve the model performance achieved in the first phase of our CIFE Seed project.
- Investigate edge cases where our approach arrives at erroneous classifications.
Relevant Links for this Research
This research will be continued and the Final Report will be issued as part of the ongoing study:
Assessing building-level retrofit potential instantaneously and on a large scale (continuation)