Part Two: How To Bathe a Dinosaur – Why Analytics Could Safeguard 90% of the Success of Iot
In Part One of this blog, I shared how McKinsey research proved that the current hype around the Internet of Things is probably understated, since the full economic impact of IoT by 2025 is valued at $3.9 -$11.1 trillion per year. To understand where the total value potential of the IoT lies, I introduced the concept of settings. Specifically, there are nine settings (like Home, Offices, and Factories) that give us a cross-sector view of that total value potential. I discussed the settings in detail last week, and today I want to discuss the two big types of opportunities driven by IoT applications—the transformation of business processes, and new business models driven by interoperability.
Transform Business Processes
Most of the collected IoT data today is not used or even monitored. The data that is used is primarily focused on real-time control or anomaly detection. As such, the first opportunity is to start monitoring the data in “its broader context” and transfer this into insights. The IoT sensor data can be used to optimize production processes. One of the applicable areas of improvement is preventive maintenance based on predictive analytics.
To amplify, the main challenge here is to narrow down data interchange with the network onto only data that is relevant. Relevant data equals information and insight. To recognize the data that is relevant, the use of predictive analytics will be the solution.
Interoperability Drives New Business Models
Mainly driven by the interoperability of IoT applications, new business models drive economic value. For example, since IoT applications allow the manufacturer to monitor both the behavior and the usage of their equipment used by their customers, they can start offering their products “as a service.”
It will permanently change the basis of competition. As an example, we could take a compressor manufacturer that has built-in sensors in its compressors that allow them to monitor health and utilization of the compressor and its output (compressed air). They can now start offering their product “compressed air.” (Watch Kaeser Kompressoren’s customer story.)
A well-known tire manufacturer puts sensors in all of their truck tires in order to monitor coordinates, health, and condition. Instead of providing this data to their customers, they sell it back (!), offering information on preventive maintenance but also connecting their information with truck assistance companies who can offer repair services if needed.
Another way interoperable IoT can create new business models is the Logistics and Cargo industry using AIS track data of ship cargo. By comparing their own ships’ traffic progress (using predictive analytics on weather models and GPS coordinates) with those of competitors, one of the leading Cargo companies is now contacting potential clients suggesting that they can shorten delivery compared to others. They combine the AIS IoT application with air traffic IoT applications to offer clients a better deal.
These two core opportunities described above will strongly rely on business analytics if they both want to be successful. Let me sift this into pieces to answer why. It all comes down to applying the closed loop portfolio of analytics towards the Internet Of Things:
If there is one component of business analytics that affects the success of interoperable IoT, it is predictive analytics. It will play a core role in recognizing, prioritizing, and interpreting sensor data that is relevant.
- Recognizing: Predictive analytics is capable of isolating the data records of the sensor that apply to a certain business case (like records that indicate abnormal behavior).
- Prioritizing: Predictive analytics is capable of operationalizing self-learning predictive models that isolate the sensor data that requires attention.
- Interpreting: Predictive analytics is capable of ….. (fill in the blank).
Towards the interoperability aspect of IoT, predictive analytics plays a dominant role. Be aware that the correlation between the one IoT application and the other is in 90% of the cases unknown. Predictive analytics allows to see whether there is a correlation, and how to utilize that correlation towards business value.
The beauty of modern predictive models is that they are not only self-learning—the model gets stronger and better every time—but they also allow you to operationalize their models. An applied and correct predictive analytics model can be made part of the IoT application permanently.
“What we want to do is go from reporting on events to anticipating them, and ultimately changing outcomes, “ says Tonio Fraizier, SVP at DigitalGlobe “The more we can get ahead of those events, the more we can equip the people who are going to take action.” ¹
The insights that interoperable IoT application can bring, add value to the planning process and as such to the office of the chief financial officer. Smart cities might adjust their lightning schedules based on real-time crime data ensuring better security. Combined with predictive models, it allows cities to better forecast electricity costs. Manufactures benefit from utilization data of their sensored equipment to better forecast sales and repairs.
As concluded earlier, most of the IoT data today is either not used or not exploited to its full extent. Though there is always room for improvements, I believe that today’s analytics tools are equipped to interactively monitor IoT data. The reason for not exploiting its full potential is due to both a lack of using predictive techniques to decide what data is relevant, plus the “consumers” of the IoT data being not ready yet to further explore with enriched data from non-sensor sources. Monitoring, enriching, and blending IoT sensor data with non-sensor data is a phase which will get its momentum this year.
A special place in IoT applications is prepared for the alerting functionality of analytics monitoring. By using event stream processing (ESP) techniques, it’s possible to apply event-based alerting in real time. It means that sensor data can act as a trigger for follow-up events. You may think in the widest possible way when it comes to alerting. Events can trigger other processes and start a new chain of activities. But it could also trigger the run of analyses or collection of new insights. The technique of ESP is ground-breaking and will boost IoT enormously. An example of an ESP system is below.
Researchers believe the software industry still has substantial work to do to provide the architectures to allow for interoperable IoT applications. I don’t. Companies like SAP offer fully-equipped architectures to apply customer-centric IoT applications, vertical-industry applications, and fully-interoperable IoT systems.
SAP’s in-memory technique provided in a full Cloud-based environment allows for:
- Performance and scalability required
- An application platform that allows for customized application development
- Connectivity towards Big Data clusters if needed
- Seamless integration with operating systems
- The analytics platform necessary for the closed loop portfolio.
¹ Tonio Fraizier was general manager of DigitalGlobe’s Insight analytics business until the end of 2014 and is now in charge of its offerings for U.S. government agencies. “What we want to do is go from reporting on events to anticipating them, and ultimately changing outcomes,” said Fraizier. “The more we can get ahead of those events, the more we can equip the people who are going to take action.” For example, if a military special-forces team is being sent on a mission into an urban setting, “you want to be able to say who can see them there or pinpoint safe places to land” based on current information. He also cited efforts to police fishing exclusion zones in the world’s oceans and seas; because of the vast amount of geographic space that’s involved, having data analysts comb through satellite images in search of illegal vessels could take “days or weeks to get an answer.” More prosaically, real-time counting of cars in parking lots at shopping malls could provide an indicator of retail traffic, said Frazier.
Other sources: The Internet of Things: mapping the value beyond the hype, McKinsey, June 2015. IoT and Digital Transformation: a tale of four industries, IDC, March 2016