Love in the time of streaming data

Manager of Portfolio Strategy, IBM

Save the date! 2033 might see the marriage of IBM and your auto manufacturer, says a recent Economist article focused on predicting the future of the auto industry. Many sessions, demos and the appearance of both a Tesla and Corvette in the solution expo clearly lay the groundwork for such a relationship. Even five years ago this would have seemed odd. But, looking back on IBM’s 100 year history of inventing, investing and manufacturing everything from meat slicers to scales to main frame computers, the future may very well include cars. Titles such as chief innovation officer and chief technology officer are being created at places like Kia and Ford and automakers continue to have a strong presence at technology-based trade shows like InterConnect, and also the Consumer Electronics Show.     

computingsystems_blog.jpgThe reason for this partnership is that vehicles of all kinds (planes, boats, trucks, cars) are quickly becoming “big data machines.” They generate data, need a place to store data (like the cloud) and require IT applications to turn this data into actionable results (like automatic brake deployments, automatic parking and self-driving functionality). In other words, they are computing systems with wheels, wings and anchors.

Vehicles in particular are a hub of streaming data sources. By streaming data, we mean continuous flow of discrete packets of digital data. There's really no starting or stopping point to the stream—it has no beginning and no end. The feed exists before your application exists and long after your application runs. The challenge is to tap into it.

Lots of ideas for how to do this were presented at InterConnect 2015. Two notable Ignite talks, including Robert Uleman’s talk on real-time geospatial analytics in the cloud and Bryan Boyd’s talk on how to remotely control devices by analyzing streaming data in the cloud, generated a lot of discussion and creative thought. Attendees at InterConnect came away with technology to help them fuel their next big business idea, and with good reason: streaming applications have broad applicability, especially in the Internet of Things, and recent reports estimate that IoT applications could raise the level of US gross domestic product by two to five percent by 2025.

Some of the most relevant and prominent technologies featured were streaming analytics platforms. Platforms like InfoSphere Streams deliver operators to perform analytics on these data streams which can be built into continuous and dynamic streaming applications. Operators include data mining, complex event processing, video processing and more. 

One real world application was presented during the Internet of Things Keynote by Kevin Massey of Cummins Inc. Streams is used at Cummins for a telematics application. Cummins partners with telematics providers to get the data about faults issued from the engine. The value-add is a series of homegrown analytics that can advise prescriptive maintenance based on the profile of the fault codes for that engine over time. Cummins then alerts the telematics providers to alert the fleet managers. They also send the alerts to Cummins Customer Care call centers. The telematics data from the providers is ingested, validated and then enriched with reference data about the engine and where it is deployed, the customer and the fault code.

What other engagements do you see in the future? Let us know in the comments below.