Transporting analytics to the Internet of Things


IoT in TransportationWhy are so many companies across a diverse set of industries investing in and around the Internet of Things? Everywhere I go, every blog I read … I sound like my favorite band from the 80s: the Internet of Things is watching me.

In reality, it’s the reverse: I'm seeing the Internet of Things (Iot) everywhere: companies investing in sensors, networking and applications with the expectation that this investment will increase revenues, lower costs and improve profitability over the short and long term.

While the term the Internet of Things was coined in 1999 by Kevin Ashton at Procter & Gamble, the mainstream application of IoT is just getting started. As the trend has heightened, I've been evaluating the potential for IoT to support better decision making in travel and transportation.

My experience in the travel and transportation industry has always been about using analytics to support decision making. In fact, I started my career as a pricing strategy analyst at American Airlines. And now I’m fully converted. You can call me an evangelist for the IoT in transportation, especially because the potential to take data coming from the IoT, incorporate purposeful analytics and reach better decisions quickly, is significant.

This week, I am presenting as a guest of the Institute for Defense & Business at the 2015 NDTA-USTRANSCOM Fall Meeting on the convergence of IoT and analytics in the transportation industry.

I am speaking directly about why I have become so enamored with the ability of the IoT to deliver business value, especially in transportation.  It’s really quite simple: the IoT delivers value through the data that the things provide for decision-making.

This data is collected by sensors and devices on railcar components, semi-truck engines, or other elements within the transportation value chain. And it is now available on a real-time, or streaming, basis.  This provides the ability to learn from trends in that data quickly and act upon those trends within seconds.

Gone are the days where data is collected over the span of weeks or months, sent to an analyst for review, and – after another week or two of data crunching – the analyst presents a report or PowerPoint to her boss with recommendations for changes.

As Tom Davenport said:

To make the Internet of Things useful, we need an Analytics of Things.  This will mean new data management and integration approaches, and new ways to analyze streaming data continuously.

So, in order to take advantage of the streaming big data (and it is big data by every definition of the phrase) coming from the sensors, we must reconsider how we use analytics.  Remember, that I didn’t say that the analytics themselves must change.  In most cases, we can use the same analytics applied to streaming data as we used in a batch model.

What we need is a good understanding of where we apply the analytics: on the edge, at rest or in the middle. Let me explain:

  • Analytics on the edge means analysis at the specific device or sensor.
  • Analytics at rest means data pulled out of the stream and used for high-performance analytic model development
  • Analytics in the middle takes place on data as it's streaming. Some analysts have called this middle ground “the fog,” and it’s relevant because it can be a combination of the streaming data itself enriched with sitting data such that we can detect more complex events sooner.

We have now arrived at a different place in the analytic continuum.  The optimal analytics experience is a multi-stage analytics experience.  It includes continuous queries on data in motion and at the edge, with incrementally updated results.  This new process moves analytics from centralized data warehouses to edge analytics, which are closer to the occurrence of the events.

What does multi-stage analytics of IoT data look like? It happens fast (seriously, we are speaking about microseconds or msecs at this point) and at very high volumes.  It requires specific business rules that give instructions on whether to save, discard, aggregate, transform or enrich the streaming data without overloading the entire system or network.  Multi-stage analytics includes pre-determined data mining, decision making, alerting, scoring, and profiling of the data to exploit the value of the streaming data.  And, it might also include managing the data differently – creating “out of order” handling to make the data source streams understandable to the analytics and the decision-makers.

We have all the building blocks in place to exploit the value of the Internet of Things and the Analytics of Things: sensors, or assets, creating data, the communications network connecting the data and the analytic and computing applications that make use of the data flowing to and from the things.

The Internet of Things can be transformative in transportation operations: in maintenance and engineering, you will have more information sooner, which means you can predict the maintenance needs of individual assets before failures occur and proactively service assets at an opportune time when your asset is near a repair facility.  This reduces costs across your operations.  In supply chain situations, you can monitor inventory levels on a near real-time basis, develop better forecasting models and optimize this inventory, when and where you need, lowering supply costs, increasing efficiency and enhancing revenue opportunities. On the customer side of transportation, you can enhance the customer’s experience by providing real-time forecasts of arrivals and notifying them sooner if delays occur.  And, happier customers are loyal customers.

The Internet of Things opens up tremendous opportunities for transportation companies, generating significant streaming data which can be relevant for decision-making.  However, it is critical to apply the appropriate analytics to streaming data in order to derive value from that data.

Multi-stage analytics is not rocket science; it’s simply the judicious application of the right analytics at the right time in the right place to the right data, which is what you need to exploit the value of the Internet of Things.  That is why I’ve become an Internet of Things and Analytics of Things evangelist.

Learn more about how companies are bringing IoT and industrial-grade analytics together in this report from Blue Hill Research, IoT Analytics in Practice.


About Author

Lee Ann Dietz

Sr Industry Consultant

Lee Ann Dietz is the Senior Industry Consultant in the Travel & Transportation Global Practice at SAS, and has almost 25 years supporting customers in T&T with analytical solutions. Prior to joining SAS in 2012, Lee Ann held various positions with Railinc, DZone, Inc., and SAS. Lee Ann began her career with American Airlines and SABRE, after earning her MBA from the Darden Graduate School of Business at the University of Virginia and BA (Economics) degrees from Stanford University.

Back to Top