ENERGY STAR brand integrity

Using artificial intelligence and machine learning to identify ENERGY STAR brand misuses

ENERGY STAR® Canada is a voluntary partnership between the Government of Canada and industry to make high efficiency products readily available and visible to Canadians. Natural Resources Canada’s Office of Energy Efficiency (OEE) formally enrolls manufacturers, retailers and other organizations as participants in ENERGY STAR Canada, while manufacturers, retailers and public utilities help promote the brand and ensure ENERGY STAR certified products are prominent and readily available to Canadian consumers. Given the scope of this partnership, the program is vulnerable to brand misuse that can be adressed through the development of digital solutions.

Project objectives

Artificial intelligence (AI) and machine learning (ML) techniques can help identify misuses of the ENERGY STAR brand on websites and social media as well as strengthen processes and refine policy to provide consumers with better information for informed product choices.

Expected results

This project was designed to identify improper uses of ENERGY STAR brand elements, using a web scraper data aggregator and ML and AI techniques. Preliminary use of these tools yielded over 2,000 cases of ENERGY STAR brand non-compliance of varying kinds, including non-certified products being advertised as ENERGY STAR certified, misuse of brand elements on product or website and non-authorized companies using brand elements.

The ML and AI techniques also identified problems with how ENERGY STAR Participants identified themselves online (such as providing out-of-date information, product certification or partner status) and offered a critically valuable insight into the disconnect between online retailers’ use of product model numbers and those of the program.

The significant advantages of AI and ML over traditional brand compliance methods in terms of speed and efficiency have already led the program to further invest in tool development. Further, it is expected that the findings can also be applied to similar compliance issues in connection with Energy Efficiency Regulations or other Government of Canada labelling or certification programs.