Open for Data: Takeaways from IBM InterConnect 2016

Post Comment
Big Data Evangelist, IBM

IBM InterConnect 2016 featured a most comprehensive set of IBM announcements—both prior to and during the conference—in regard to its deep investments in cloud data services, open analytics, cognitive computing and the Internet of Things. Each of these moves are advancing IBM’s Open for Data strategy, and I came away from IBM InterConnect 2016 with other takeaways and some thoughts I had while attending various sessions throughout the week.

Open cognitive Internet of Things

The Internet of Things, in my view, was the most innovative focus of the IBM InterConnect 2016 sessions. A noteworthy innovation discussed was the convergence of cognitive computing with the Internet of Things—also known as cognitive Internet of Things. Much of this discussion centered on the disruptive potential of open platforms, tools and programmatic interfaces that enable developers to produce a staggering assortment of innovative products and services that can impact our lives.

For a great customer example of cognitive Internet of Things, stream the first day’s general session in which Matthias Rebellius, CEO at Siemens Building Technologies, discusses IBM Watson Internet of Things as an enabler for intelligent management of connected sensor-laden buildings. Also, take a deep dive on how IBM is approaching cognitive Internet of Things by streaming the replay of the thought-leadership session, “Cognitive Internet of Things: The New Leadership Agenda” on IBMGo.

In addition, you can stream Harriet Green’s talk during the second day’s general session to hear an IBM executive’s perspective on what IBM is doing on the investment and partnership front to advance cognitive Internet of Things. Green, general manager for IBM Watson Internet of Things at IBM, described cognitive Internet of Things as “Watson plus Internet of Things.” She underlined IBM’s Internet of Things leadership, including the more than 4,000 clients, 1,700 partners, 10,000 security clients, and 30 industry solutions. Green also honed in on the significant new IBM partnership in the open source Internet of Things hardware arena with a clear focus on cognitive Internet of Things.

In particular, Green discussed the work IBM is doing with the Raspberry Pi Foundation to enable Watson cognitive analytics to execute on Internet of Things endpoints that incorporate that group’s open source hardware technology. As Green discussed this effort, I immediately saw its relevance to some other cognitive Internet of Things and open source announcements that IBM made in February 2016 and at IBM InterConnect 2016.


Prior to IBM InterConnect 2016, IBM announced it is submitting Quarks, a development tool for embedding in-memory, cognitive analytics software in myriad Internet of Things platforms, to the Apache Software Foundation. As I discussed in a blog post, Quarks can manage and analyze continuous streaming data on any Internet of Things–capable device. It provides a single runtime for analyzing Internet of Things data at the edge, it can run on Internet of Things–capable edge devices and gateways, and it enables continuous correlation of data across the Internet of Things. Quarks, now available at GitHub, works with Apache Spark, Apache Hadoop and many other data and streaming analytics environments.


At IBM InterConnect 2016, IBM announced that it is unveiling an open source software tool for event-driven programming of data analytics, such as cognitive Internet of Things, as composable microservices. The newly announced OpenWhisk enables developers to build feature-rich, event-driven applications quickly and easily. By open sourcing this technology, IBM intends to generate a powerful ecosystem of event providers and consumers to develop the platform. Developers can use OpenWhisk to quickly build microservices that automatically execute cognitive analytics and other Internet of Things software code at the Internet of Things–capable edge device in response to events such as receiving sensor data.

Developer ecosystems

IBM and GitHub announced at IBM InterConnect 2016 an open source repository of more than 140 application programming interfaces (APIs) and services for cognitive Internet of Things application development, as well as new sources of data. The initiative is expected to enable developers to view, collaborate, integrate APIs, iterate, deploy and update on a single platform. The significance of this effort is that GitHub, the largest host of source code in the world, offers complimentary repository services that are used in numerous open source software collaborations.

Open for Data

I closed out my experience at IBM InterConnect 2016 by attending the fascinating panel on IBM’s open for data strategy, “How Open Analytics Is Changing Cloud App Development.” Benjamin Tao, director of worldwide portfolio marketing in IBM Analytics Platform Services, moderated the panel’s discussion. The panel included invited experts Steve Ardire, Adrian Bowles, Bob Hayes and Roger Strukhoff, and the discussion covered many topics. Although the discussion didn’t touch directly on cognitive Internet of Things, the topics were germane to cloud data services in general, including but not limited to cognitive Internet of Things.

Panelist Kamille Nixon, senior portfolio marketing manager in cloud data services at IBM, remarked how users have come to expect intelligent Internet of Things and other applications for real-time predictions. And based on those insights, they expect to adjust their responses and recommendations dynamically. In my mind, Nixon suggested cognitive Internet of Things when she stated, “people have this expectation that they’ll be able to have all these functions in all apps.” Nixon then listed the specific functions she had in mind: “I’m convinced that most apps will have full text search, as well as embedded recommendation engines and embedded graph databases.”

Nixon’s comments during the panel discussion amplified what I heard Adam Kocoloski, CTO, cloud data services at IBM, state during a thought-leadership session. Kocoloski, while also discussing recommendation engines as a key application for graph analytics, pointed out the wide range of problems this approach is well suited to resolve: “Sometimes people don’t know they’re dealing with a graph-shaped problem.”

Although I didn’t hear anybody at IBM InterConnect 2016 explicitly state that cognitive Internet of Things is such a problem, I am now quite convinced that it is. Cognitive Internet of Things, which finds meaningful patterns in unstructured sensor data, would seem to be the perfect use case for graph modeling.

Over lunch one day at the event, Nixon and I discussed the notion of what exactly such a problem might look like. A graph-shaped problem is one that is best represented in the database by networks of nodes and edges, rather than relational tables and primary key relationships. Autonomous vehicles, for example, generate ample sensor data, as do geospatial applications on smartphones and other dynamic Internet of Things–capable edge devices. To stay contextually and predictively oriented in complex environments, these applications depend on cognitive Internet of Things capabilities, such as those built in Quarks, to make sense of the unstructured sensor data that describes their changing environments. Recommendation engines enter the picture as the graph-powered, next-best-action engines for ensuring continual optimization of Internet of Things endpoint actions in increasingly complex, real-world scenarios.

In my mind, these autonomous intelligent Internet of Things edge-node scenarios practically scream for cognitive Internet of Things. And those markets can’t truly come into their own as robust ecosystems without a deep stack of open source data, analytics, hardware and other components.

Get your take

These are my personal takeaways from IBM InterConnect 2016. Take your own personal open for data journey to the next step. And to review Open for Data, Cloud Data Services and cognitive Internet of Things content from IBM InterConnect 2016 sessions, check out the event’s website, Tumblr page, Twitter page and Facebook page.