By choosing AWS, we increased development speed by 62% for the WiZ infrastructure and avoided an on-premises solution.
Jerry Yuen DevOps Lead

Founded in 2004 and headquartered in Hong Kong, TAOlight (TAO) offers customers the technology to manage their lights and lamps. The company produces light systems that are designed in Paris and manufactured in Shanghai, and employs around 20 software developers, firmware, and backend engineers in Hong Kong. TAO, which has manufactured more than 80 million lights since it launched, offers its two major brands—iDual, a smart light-emitting diode (LED) remote-controlled lighting system, and WiZ, a smart system which offers different ways of controlling lights within rooms over Wi-Fi or remotely through the cloud using an iOS or Android mobile app.

Since its launch, TAO has taken an innovative approach to lighting. In 2009, it became one of the first companies to manufacture LED bulbs for the European market. LED bulbs are highly efficient, converting as much as 95 percent of their energy into light and only five percent into heat. Following the success of its LED bulbs, TAO began designing lighting systems that incorporated technology. For example, in 2014, TAO introduced iDual with its own infrared remote controls, and in 2016, TAO began designing its WiZ product range. The company looked to integrate Internet of Things (IoT) technology into WiZ as it wanted customers to be able to adjust light settings using an app on their smartphones, and eventually using voice commands to control lighting.

TAO needed a strong IT infrastructure for its backend systems to support WiZ. “We wanted a platform that could drive innovation and enable us to bring IoT technology into our designs,” says Robert Chen, brand director at TAO. The company also wanted a cloud infrastructure because it would help distinguish WiZ from other lighting brands. Chen explains, “App-based controls for other light systems store and process data on the device running on the app or an external hub. We wanted processing done in the cloud so if a device was lost, customer data wasn’t lost along with it.” TAO also wanted to be ready to incorporate big data and analytics so designers could consult usage data when designing TAO solutions in the future.

From the start, the company rejected the idea of an on-premises infrastructure to support WiZ. Jerry Yuen, devops lead at TAO, says, “With a cloud architecture, we could reduce infrastructure development time and labor costs.” For Yuen, it was important to find a cloud provider whose solutions could help lower IT production costs. “We are always looking to reduce application costs as part of a broader goal to increase IT efficiency and deliver more value back to the company,” he says.

TAO chose to engage with Amazon Web Services (AWS) after assessing cloud services from other leading providers, including the Google Cloud Platform and Microsoft Azure. Says Yuen, “We felt that AWS provided a wider breadth of services for our current and future needs. We did some calculations and found that AWS was more cost-effective in the short term and long term.”

Members of the IT team already had experience working with the AWS Cloud. The only gap in their knowledge was around building an IoT architecture in the cloud. Yuen says, “With IoT architectures, the connection between your devices and the backend infrastructure has to be maintained at all times. AWS helped us create a design that ensured connectivity while maintaining a high level of security. Of all the IT solution providers that I’ve worked with, I received the best support from AWS.”

The WiZ backend infrastructure sits inside an Amazon Virtual Private Cloud (Amazon VPC) for security. Data entering the infrastructure from devices hits proxy servers running on Amazon Elastic Compute Cloud (Amazon EC2) instances. Behind the proxy servers sit the main application servers—also running on Amazon EC2—which run Amazon ElastiCache to enhance application performance. The infrastructure database, which defines the relationship between customers and their WiZ products, is provided by Amazon RDS for Aurora (Amazon Aurora). Finally, the infrastructure features Amazon QuickSight, which provides TAO personnel with performance analytics and business insight from devices and IoT lighting.

TAO now ships WiZ products which are compatible with Amazon Echo, a hands-free speaker that is voice-controlled. Amazon Echo is integrated with the WiZ infrastructure through the Alexa Voice Service (AVS) so that customers can control the lights in their homes and offices with voice commands. AWS Lambda transfers the voice request to the infrastructure’s application programming interface (API) server, which in turn triggers the events for the light settings to be changed.

By using AWS, TAO has achieved its goals while avoiding the IT challenges associated with on-premises infrastructures. The company has successfully delivered an IoT-enabled lighting solution to market. What’s more, TAO has integrated app and voice-based commands into WiZ to make lighting control more convenient. Chen says, “We have distinguished WiZ from other lighting brands through cloud-based IoT technology with AWS. Because the WiZ infrastructure runs in the AWS Cloud, customers can remotely control their lights through any smartphone. All they need is an internet connection.”

TAO has incorporated analytics into WiZ using Amazon QuickSight and plans to build a big-data platform in the future as the WiZ product line grows and its customer base increases. “With Amazon QuickSight, we are gaining insight into how WiZ lighting is being used by our customers. The information will help us to improve the designs of our WiZ products.”

The company reduced the time needed to develop the infrastructure behind WiZ by adopting the AWS Cloud. “We increased the speed of development for the WiZ infrastructure by 62 percent by choosing AWS and avoiding an on-premises solution,” says Yuen. As a result, TAO has also lowered labor costs for developing the WiZ solution.

TAO is also achieving its goals of lowering IT production costs with AWS. Recently, the TAO IT team began using t2.medium instances for Amazon EC2 when using Amazon Aurora. The t2.medium instances provide a more cost-effective option for smaller workloads than larger Amazon EC2 instances. Says Yuen, “Through the introduction of t2.medium instances to support Amazon Aurora, we’ve cut our IT production costs for WiZ by 60 percent, enabling IT to deliver greater value back to the wider TAO business.”

Learn more about AWS Internet of Things.