Internet of Things (IoT) has been a hot conference topic for several years. The most quoted prediction from Gartner is that there will be 26 billion IoT devices connected worldwide by 2020, generating US$300 billion of revenue in related products and services.
IoT is not just hype and will soon make an impact on Hong Kong’s IT community because of the following factors:
Smart city initiatives -- According to a study published by Navigant research and written by the founder of DDJ Consulting Donald Johnson, smart city market size in the Mainland will grow from US$1 billion in 2014 to over US$5 billion in 2023.
Aiming to harness technology and connected devices to improve citizens’ lives, the governments of Hong Kong and other Mainland cities are rushing to launch smart city initiatives to stay competitive.
In the 2015 Policy Address, the Hong Kong SAR government designated Kowloon East as a pilot area to explore the feasibility of developing a smart city. In April this year, Hong Kong’s Smart City Consortium signed an agreement with Mainland’s Smart City Development Alliance to jointly promote business exchange.
Adoption of cloud based services -- Adoption of cloud infrastructure services is growing at unprecedented speed. According to Synergy Research, worldwide revenue from cloud services grew 52% in 2015.
The major cloud service providers are providing highly scalable, on-demand services for big data analytics, such as Amazon Web Services’ EMR and Microsoft Azure’s HDInsight. To a large extent, such services lower the entry barrier for many organizations wanting to collect, store and analyze large volumes of data generated by connected devices.
Big data getting more mature -- Collecting data from connected devices is only the first step in IoT. The implementation will succeed only with knowledge of how to analyze and make the best use of the data to identify actionable items and to make predictive analysis that bring real business value. With recent advances in big data analytics tools, technical-savvy business analysts are becoming more competent in performing the job that used to require hard-to-find data scientist talent.
The number of devices and frequency of data collection from IoT deployments are creating a massive volume of data. To put it in perspective, Cisco in its Global Cloud Index report estimated that a smart city of one million would generate 180 million gigabytes of data per day by 2019.
The collection, transmission and processing of this huge quantity of data pose new security, capacity and connectivity challenges to IT professionals. Although cloud infrastructure is popular for big data analytics, transferring the entirety of such huge quantities of data to a single location for processing is not always technically and economically viable.
DCs at the edge
In order to address these challenges brought about by IoT, more data centers are expected to move to the “edge”, locating closer to the data source. Organizations will find it more cost-efficient and technically feasible to aggregate data in multiple distributed data centers for initial processing. Only relevant data will eventually be forwarded to a central facility for further processing.
This approach of moving workloads to the edge nodes closer to the producers or consumers of data is not new. We have seen similar phenomena for OTT contents and mobile data communications. These edge nodes will require excellent connectivity to major cloud infrastructure providers and network service providers Carrier-dense data centers and those having direct and secure access to the cloud (via AWS’s Direct Connect, Azure’s ExpressRoute and the like) will have a definite competitive advantage in the coming era of IoT.
On the other hand, we should also expect some cloud infrastructure service providers to start building smaller local data centers that are closer to the data sources and consumers to provide the capacity needed.
Peter Yan is a life-time fellow and Honorary Advisor of the Hong Kong Computer Society. He is also the executive director and CEO of SUNeVision.