Over the past year, I’ve noticed that more and more business people are using the term “data analytics” rather than the more dramatic (and reductive) “big data” to refer to the collection and correlation of vast amounts of information for business.
I am certain that the adoption of the more ornate but accurate term indicates a general shift in focus from the notion of a mound of potentially untamable data to a more sophisticated appreciation of the value that rational, stepwise analysis can unlock from the information.
A recent Unisys poll of more than 100 IT professionals certainly reinforces the mainstreaming of data analytics as an essential tool for business advantage. We asked respondents about their organizations’ plans to use data analytics in support of digital services – using all components of the IT infrastructure connected for fast, always-on service delivery. Analytics-based service management, which personalizes delivery of IT and business services and pre-empts service outages, is an increasingly crucial means to streamline and guide digital business.
We found that 45 percent of respondents to the question – the largest single group – have plans in place, while another 24 percent are excited about the potential payoff from data analytics.
I also have concluded that the mainstreaming of data analytics will continue in 2016 – leading, like many technology phenomena entering a phase of maturity – to important but essentially incremental, non-disruptive changes in the landscape. But as that steady acceptance continues, progress toward more remarkable applications will begin under the radar as well.
I predict the following developments in data analytics for 2016:
- In-memory analytics platforms such as SAP HANA and Apache Spark will continue to grow in popularity – not only because of their power, but also because of the operational efficiencies and lower costs they generate by reducing data traffic cross the I/O infrastructure.
- Apache Parquet will supersede Apache ORC as the preferred tool for managing columnar-stored data in the Hive on top of Hadoop. While ORC is a valuable tool, Parquet has an advantage through its superior ease of encryption. That capability will prove to be a major edge as the role and value of analytics expands in intelligence, national security, and investigative applications – especially with increasing use of micro-segmentation to keep sensitive data within discrete communities of similarly permissioned users throughout a commercial enterprise or government entity.
- Cloud alternatives will proliferate. Many enterprises in certain industries, such as financial services and healthcare, as well as many public-sector agencies, hesitate to put sensitive information related to their business and customers in the cloud. Those users will begin to gravitate toward “analytics in a box” solutions that center around a Hadoop cluster comprising a secure server, analytics software, and related resources. They can develop analytical and business applications using the sensitive data in a private, secure environment.
- The Internet of Things (IoT) will continue to morph into the Internet of Data. I’ve seen estimates of devices connected to the IoT by 2020 ranging from 25 to 75 billion (with the count more likely to be somewhere in the middle), and as many as 1 trillion by 2025. Collecting and correlating information from even a fraction of those devices will become a major challenge. I believe that analytics leaders will accelerate efforts to create solutions that address this impending data tsunami.
- Solutions for performing analytics on unstructured data will proliferate. According to Forrester Research, analytics solutions for unstructured data lag those for structured data by as much as 30 percent. That gap surely will close as commercial and governmental needs intensify, especially those driven by requirements such as national security, forensic investigations, and law-enforcement accountability. Data scientists and software developers will need to find new ways to integrate and analyze data from a wide and growing, range of media: CCTV/IP, body, and cell phone cameras; audio streams; and e-mail messages and tweets; to name just a few. Law-enforcement applications add an extra sensitivity: Storage and analysis must often be conducted without breaking the chain of custody critical to legal evidence.
- Work will accelerate using data analytics as the catalyst for the evolution of machine learning into cognitive analysis and artificial intelligence. Increasingly sophisticated solutions drawing on technologies such as Microsoft Cortana personal assistant will enable robots to analyze data so rapidly that they’ll be doing true decision making. For example, these sophisticated capabilities will drive the transformation of services such as Teladoc, which relies on interaction with human doctors, into a fully automated service driven by software-based machines making actual medical decisions based on patient input analyzed and applied in seconds.
On the whole, I believe that 2016 will be a year in which data analytics continues to plateau in terms of mainstream business acceptance, but also begins to evolve into a basis for exciting new solutions that, sooner rather than later, will make what we used to consider science fiction a reality.
Dr. Rod Fontecilla is Vice President of Application Services for Unisys Federal Systems. In this role, Dr. Fontecilla leads all aspects of software development, system integration, mobile development, and data analytics focused on the federal government. He also leads Big Data Analytics (BDA) globally for all of Unisys. In this capacity, he is responsible for creating data analytics products for multiple industries creating predictive models using machine learning algorithms. He leads a cadre of data science professionals providing business insights to our customers. Dr. Fontecilla is responsible for providing leadership, coordination, and oversight on all IT solutions, data analytics, emerging technologies, and IT services delivery. He has more than 25 years of professional experience in the implementation of large and complex mission-critical IT solutions for both commercial and the federal government. Dr. Fontecilla brings an extensive background and expertise in data analytics, cloud computing, mobile development, enterprise architecture, and IT governance and strategy.
Subscribe to Data Informed for the latest information and news on big data and analytics for the enterprise.