R&D Focus

BESTAI, an Energy-Saving Platform for Buildings

Video of BESTAI.

Buildings consume one third of the world’s energy, of which private homes and commercial buildings constitute the majority. However, promoting energy efficiency renovation for small commercial structures as well as residential buildings has been difficult for several reasons. Firstly, owners of small structures normally lack energy expertise. Secondly, small budgets cannot afford the energy diagnostic cost that involves site visits by energy experts and long-term measurements. Finally, energy service companies are not interested in small projects unless a favorable return on investment (ROI) can be proven before the project diagnosis starts, a catch-22 dilemma. To cope with these difficulties, ITRI presents BESTAI (Building Energy Simulation Technology with Artificial Intelligence) to allow non-specialists to diagnose buildings and obtain high ROI energy efficiency investment.

BESTAI is an energy-saving expert system with a user-friendly interface that allows non-specialists to easily perform energy diagnosis for commercial buildings and residential buildings. It provides energy-saving strategies, ROI analysis, and real-time integration of geographical information systems (GIS) to analyze power consumption of buildings across multiple geographic zones. With BESTAI, users are able to conduct precise and low-cost energy-saving diagnosis for a large number of buildings in a short period of time.

Based upon computer simulation of building energy consumption, BESTAI software operates on a cloud-based platform and features an extensive database of standard building energy models (SBEMs) for building data entry. Moreover, a complete and updated regional equipment and building materials database was established using web crawler technology to automatically update equipment and material data from reliable internet resources (e.g. Energy Label and Energy Star product information), thus improving the precision of the building energy model. For diagnosis, the software can quickly rank power consumption of equipment, and then search the database for alternatives to the highest power consuming equipment. With the automatic evaluation scheme, the program will also calculate the electrical savings and ROI period for each equipment option, and recommend users the best energy-saving strategy. A large volume of completed building performance analysis data is stored and integrated into real-time GIS, enabling large-scale building energy consumption analysis. BESTAI’s built-in interface can also graft onto the current energy management system and collect data to update building models automatically in real time.

Building performance database integrated with GIS.

Building performance database integrated with GIS.

Traditional building energy simulation is a time-consuming and error-prone task, especially the process of establishing building energy models, which requires long-term professional dedication to manually input hundreds to thousands of parameters. Table 1 shows the comparison between the traditional method and BESTAI in terms of diagnostic cost, diagnostic time, and energy-saving performance. BESTAI uses SBEMs, equipment and building material database, and a user-friendly graphical interface to enable general users to quickly construct a building model and conduct building simulation and analysis for a large number of buildings, such as chain stores. There is currently no other known software that can achieve this.

Table 1. Building Energy Diagnostic Method Comparison.
  Traditional Diagnostic Method BESTAI
Diagnostic Cost (for a Typical Convenience Store)
  • NT$30,000/shop including site inspection, measurement, and evaluation by professional.
  • Almost no cost except 30 minutes labor cost of a nonprofessional without site visit.
Diagnostic Time
  • Single building: up to 1 week by a professional
  • 100 buildings: up to 2 years by a professional
  • City scale (e.g., 1,500 convenience stores in Taipei): almost impossible due to time and cost restrictions
  • Single building: 30 minutes by a non-professional
  • 100 buildings: 1 week by a non-professional
  • City scale (e.g., 1,500 convenience stores in Taipei): within months
Energy-Saving Results and Analysis Error
  • Because professionals have different levels of experience, different technicians may produce vastly different energy- saving results.
  • Frequent occurrence of low-efficiency or excessive designs, with 20-50% energy-saving evaluation error.
  • Uses systematic considerations to conduct energy savings and ROI analysis, providing optimal and objective suggestions.
  • Uses a machine learning method to automatically update and calibrate building models, resulting in an energy saving evaluation error of around 5%, greatly exceeding ASHRAE (American Society of Heating, Refrigerating and Air-Conditioning Engineers) international standards.

BESTAI has been verified in demonstration in a number of convenience stores, bank branches, restaurants, small offices, and other commercial buildings in Taiwan. The average energy-saving rate is around 15%. In our pilot project working with the Hua Nan Commercial Bank, BESTAI has achieved the saving of 4.14 million KWh of electricity annually for the bank’s nearly 200 office branches. FamilyMart, another BESTAI user, was also able to analyze hundreds of convenience stores for energy efficiency measures within a couple of weeks. In summary, BESTAI can bring good ROIs to small and medium business owners, who used to be the last group of people to consider energy saving for their buildings.