Daikin Malaysia, a part of the Japan-based air conditioner manufacturing giant, is targeting machine learning capabilities to improve customer experience following earlier improvements to the backend using Amazon Web Services (AWS).
Daikin’s senior manager for platform development and business improvement, Teoh Kuang Yee, said at the AWS Innovate event they are using AWS IoT services to harness data from their air conditioners and deliver value for both their businesses and consumers.
The firm is exploring machine learning with Amazon SageMaker and data analytics via managed cluster platform service Amazon Elastic MapReduce (EMR) to improve total customer experience and make better-quality and efficient user-friendly products.
Teoh added that the decision came in when their teams were quite disappointed with customer feedback about poor remote control response and limited resources that could integrate virtual assistants with existing wi-fi products.
Daikin had looked to switch from conventional servers to a new IoT solution that could start small and scale fast with fast response, thus leveraging the serverless computing platform AWS Lambda.
With the AWS Shared Responsibility Model, Daikin has leveraged a combination of AWS Lambda and Amazon Cognito, allowing it to focus on core functions and resource configuration.
Unlike conventional hardware, Teoh said IoT solutions had helped Daikin to easily innovate new "smart" products and services.
“We could set out more value-added services to our customers, 10X times faster than before,” he said.
Go Daikin app
With AWS, Daikin has built a mobile application which had started just as a remote control function that later covered various aspects of the customer journey.
The app - Go Daikin - is a marketing module that assists customers in pre-sales introducing their products, promotions or offers and post-sales services like maintenance booking, energy usage monitor and control, smart control module for remote access, integration with voice assistants and more.
It has also included a heat-load calculator to help customers in sizing and product selection.
“Connected products with AWS IoT core give convenience to the customers in monitoring and controlling products from anywhere,” he said.
On the operational front, Teoh said they focussed on the idea of being "fully serverless" for customer-facing services and AWS Cloud allowed seamless auto-scaling, and "worry-free" operating system security and maintenance.
They had implemented an architecture with services like AWS Lambda, Amazon DynamoDBfor static data, AWS IoT device defender for enhanced security, AWS IoT device management, Amazon Cognito for user authentication and access control and so on for real-time data analysis and all other computational needs.
Daikin also uses CloudFormation to deploy and update the production cloud after fully testing a new code.
“This restricts developers' direct access to the production cloud and eliminates the chance of accidents,” he added.
Moreover, the firm was also able to quickly build a content management system for internal marketing staff usage with existing tools and codes, storing the output automatically in Amazon S3 for a better customer experience, said Teoh.