Through the Looking Glass: Critical Asset Insight and Transparency Increases Operational Efficiencies & Customer Confidence

A looking glass is a magical lens or portal through which things can be seen that are otherwise invisible. This is a perfect metaphor for the increasing challenge faced by businesses to find value hidden in the data they generate as well as data that they have access to. Value nuggets in data are often obscured by large volumes of data ("noise", if you will). The larger the data sets, the more obscured the value nuggets are. A theoretical looking glass capable of discerning the signal from the noise will find value nuggets that enable precise and timely reaction to situations, as well as proactive and prescriptive measures that result in real quantifiable benefits.

Erik Brynjolfsson[1] characterizes the benefits as a "data payoff." Results of his study of 179 large companies revealed that companies that used data-driven decision-making achieved productivity gains of up to 6%. Brynjolfsson further asserts that "a 5% increase in output and productivity is significant enough to separate winners from losers in most industries."

Seeing this payoff, more and more businesses are adopting a looking glass approach to understand the status of their operations and make on-the-fly positive improvements. Information and insights extracted from all data at rest and in motion enable new levels of efficiency for operations of all types and across broad geographic regions. Examples include distribution pipelines for water, oil and gas; global manufacturing operations; and transportation and logistics businesses.

Beyond internal efficiencies, businesses must also satisfy consumers who are becoming accustomed to transparency and have ever-increasing expectations of the companies they choose to support and do business with. Consumers are also accustomed to real-time responses and will overwhelmingly choose providers that deliver relevant information that is an integral aspect of an overall outstanding user experience.

Similarities between the Challenges and the Solutions to Big Data
The growth of data generation is due to a confluence of factors - Moore's Law and Metcalfe's Law. Powerful sensors embedded in many devices, including smartphones, tablets, and wearables[2], are proliferating because of declining size, lower cost, and increasing processing power, all attributable to Moore's Law. The ability of devices of all types to communicate through ubiquitous networks and make the information they capture and transmit available for analysis and storage is what Robert Metcalfe postulated in the early 1980s.

The same laws are at play regarding a looking glass solution, especially Moore's Law which is at the foundation of continuously increasing processing power that enables sophisticated in-memory analytics capable of executing in real-time. Continuous advancements in servers, storage, and software are the foundation for a looking glass, regardless of whether it is labeled as "artificial intelligence," "machine learning," or simply "analytics."

Impact of the Internet of Things and Big Data
Ongoing proliferation of low cost, battery-powered, sensor-equipped devices and ubiquitous communications are resulting in opportunities for businesses to peer through their own looking glass and see hidden value in their data that they can transform into actionable insights using advanced analytics. Businesses must peer through a looking glass to find insight hidden in all of their data.

Data volumes are increasing rapidly, especially data generated from sensors. The quantity and velocity of data generated is so great that not all of it is stored or analyzed. You can envision streams of data like water containing gold nuggets flowing into an ocean; data that is not analyzed while flowing and/or diverted for storage and subsequent analysis to extract the nuggets is lost forever. It is therefore imperative that data be analyzed while it is flowing so that you can now see your data, in motion, in real-time, to more deeply understand your situation in context.

This approach to analyzing data for insights is called situational intelligence and it lets you act quickly and confidently by providing a view of every situation from multiple perspectives. Situational intelligence also provides prescriptive suggestions and remedies, enabling you to proactively make beneficial decisions such as preventative maintenance.

When people think of big data they generally think of ecommerce and media properties: Amazon, CNN.com, Facebook and Google, to name a few. Such businesses capture every aspect of electronic end-user interactions with their properties.

A lesser-known source of big data is generated by electronic sensors that monitor the status of organizations' assets and operations. Manufacturing, mining, energy generation and/or distribution businesses, as examples, generate and capture massive amounts of data from their business operations. These industries and businesses capture data from physical assets, many of which are "smart," which is to say they have embedded sensors and are able to transmit telemetry data. Smart devices are replacing "dumb" devices; as an example many power utilities are replacing their usage meters with smart meters, obviating human meter readers from making recurring visits to read and record usage. In addition, new types of smart devices are being invented and deployed. Transportation and logistics providers generate and capture massive amounts of data from their business operations, especially in-vehicle telematics. Data from such devices is growing at an increasing rate, and much of it is neither captured nor analyzed.

Within the business and industrial sectors such smart devices are being referred to as an Internet of Things (IoT), or connected, devices. As noted above, organizations must embrace not only IoT, but the ability to harness the valuable information and insight they provide. Several interrelated technologies are required to derive not just insight, but at-a-glance actionable insight; key among those technologies are analytics capable of operating on very large data sets (aka big data) in real-time.

Increasing Your Operational Efficiency and Productivity
The challenge that businesses face is capturing, aggregating, and analyzing their data to find patterns, trends, clusters, anomalies - insights, if you will - that never before would have been found. Taking this a step further is to make those insights readily visible to the appropriate decision makers and/or to other systems and processes for real-time action. Greater throughput is an obvious advantage of widening or removing bottlenecks, especially in the case of automated machine-to-machine decision making. Another operational advantage and productivity enhancement is that your managers and staff will have at-a-glance assurance that everything is optimal and okay, and when that's not the case alarms and alerts will direct their attention accordingly. Depending upon the situation and the capabilities of the analytics, recommended actions and remedies may also be provided by the analytics solution.

A transformation of your business to this operating state and tempo will position your business for future success and avert competitive defeat or overall obsolescence. Benefits include streamlined internal processes, more productive field workers, the detection of unauthorized or rogue use of your resources, and greater availability of your plant, network(s), and physical assets. An example of this is predictive insights, which drives proactive maintenance to increase overall uptime that in turn assures production capacity and compliance with service level agreements and the like.

Consider the case of a large transportation logistics company. Knowing the exact location and use of its assets in the field will save millions of dollars each year. This company's field assets are taxed differently when they are on-road versus off-road, so having precise location and time-of-use information reliably streamed from in-vehicle telematics (without human errors) is essential. Analyzing that information enables reporting with never-before-possible granularity that eliminates rounding assumptions of the past, lowering their operating costs.

Enhancing Your Customers' Experiences
Bringing relevant real-time information forward in readily digestible formats to your end users and other stakeholders gives your business many opportunities to differentiate itself and realize your competitive advantage.

Imagine two airport shuttle services, one with in-vehicle telematics that provides current location and temperature inside the vehicle and one that does not. End users are more apt to choose a vehicle where they have a high confidence in an exact pickup time that is calculated based on the distance of the vehicle to their location and current traffic. The shuttle information is even more compelling and of greater impact to the selection of vendors if the end user is aware that the interior of the vehicle is a comfortable 72 degrees. As this example highlights, the user experience and interactions will increasingly include real-time information and insights from an organization's physical assets.

Another common example is parcel shipment and delivery. Consumers want to know when their purchased items or parcels have shipped, where they are now, and a reliable estimated time of arrival with as much granularity as possible (i.e., to within an hour). Such information and transparency between consumers and businesses is increasingly important to attract and retain consumers who have an ever-increasing palate of options to purchase and receive the items they need and want. If you think about it, sensors and ubiquitous communications make it possible to connect a parcel to a conveyance and to the person awaiting delivery.

Implementing Your Looking Glass
I am clearly an advocate of using data analytics to drive high-confidence business decisions. Moving in this direction generally impacts your entire enterprise, so I recommend a phased approach and the following steps to putting in place your own metaphorical looking glass.

As a first step I recommend your departmental leaders, including your IT team, be involved early in the process. The next step is to establish a vision of how your business will operate after becoming proficient at data-driven decision making. The vision should include explicit and measurable goals and corresponding use cases. If appropriate to your business operations your long-term vision needs to identify whether the ultimate goal is analytics-aided decision-making and/or fully automated decision-making (e.g., machine-to-machine decision-making).

Another critical early step is to identify all internal data sources and additional complementary external data sources that will be needed to support your decisions and goals. To ensure that your analytics program moves forward with little or no unanticipated delays, you will also need to assess data quality and data accessibility. An experienced analytics vendor can work with you to assure successful integration of data into your analytics program.

In addition, make sure the vendor and/or system integrators you choose have the capabilities and a track record of delivering and successfully implementing reliable and scalable enterprise analytics programs. Important capabilities to assess and require are: capabilities of the core product, extensibility of the analytics, professional services that includes data science, training, and support. If you anticipate extending your analytics program to multiple use cases, you should give strong consideration to a platform. In the case of an analytics platform, you should also favor a broad pallet of algorithms, the ability to easily add custom analytics, and a supporting ecosystem of plugins, developers and integrators.

To maximize early successes, I recommend to constrain your analytics project by choosing one or a small number of realistic and achievable goals and align the early phases of your implementation to what is necessary to achieve those goals. Build on your successes and momentum by incorporating more goals and use cases into your analytics program. Also make sure to review the program, including the vision, goals, phases, successes, lessons learned, and areas for improvement as widely as possible.

Summary
Value nuggets hidden in your data truly deliver a quantifiable "data payoff" too valuable to forego. Lagging or failing to find value nuggets and not using analytics to facilitate data-driven decision-making will place your company at greater risk of competitive disadvantage or obsolescence.

Situational intelligence empowers your staff and your company to act quickly, decisively, and confidently in any situation. Realizable benefits include increased productivity, increased customer satisfaction, and competitive advantage.

References

1. Erik Brynjolfsson is the Schussel Family Professor at the MIT Sloan School of Management, Director of the MIT Initiative on the Digital Economy, Research Associate at NBER, and Chairman of the MIT Sloan Management Review. His research examines the effects of information technologies on business strategy, productivity and performance, Internet commerce, pricing models and intangible assets.
2. Smartphones, tablets, and wearables typically include several of the following: an audio sensor (a microphone), a video sensor (a camera), position & motion sensors, a human pulse sensor, a temperature sensor, and a GPS location sensor.