Augmented Intelligence
Sep 16, 2016

Preparing for an IoTA setup

Internet of Things (IoT) is the most talked about and pervasive technology that is about to impact a wide variety of industries and future businesses. The purpose of this blog is to share tips on planning an IoT Analytics (IoTA) setup.

The IoT era
Internet of Things is the new highway of information generated and information collected—information which seamlessly flow from consumers to companies. We are starting to belong in a world of intelligent electronic devices connected via Internet, whether through mobile technologies, low-energy sensors, beacons, or RFID chips. The important question, however, will be, "Are we listening?"—or, more aptly, "Are we getting ready to listen?"

For example, we all are aware of the huge rush for fitness wristbands like Fitbit. Progressive Insurance's Snapshot offers lower rates in exchange of cars that share driving habits with them. Nest offers a connected home monitoring system. While Apple's mobile voice control Siri learning your likes and habits.

The research and advisory firm Gartner predicts there will be 25 billion connected "things" by 2020, compared with 4.9 billion by the end of 2015. That's 38.5% year-on-year average growth!

Future role of data insights companies in the IoT era
IoT essentially brings in a massive influx of real-time data.

Fast (speed and change), huge (volume), heterogeneous (diverse) and inconsistent (noisy).

With traditional data insight models in place, whether descriptive or predictive, we still depend on manual, or at most semi-automated processes to create these insights. Looking around at the analytic processes around us, we will find expert analysts using tools to read, sanitize, summarize, and then find patterns in that summary. Patterns that either will tell what happened or will leave pointers to imagine what can happen.

However, none of our existing systems scope up to stay sane in the evolving information vortex.

This means we need to look at our preparation for the imminent future.

Three men were laying brick. The first was asked: "What are you doing?" He answered, "Laying some brick." The second man was asked: "What are you working for?" He answered, "Five dollars a day." The third man was asked: "What are you doing?" He answered, "I am helping to build a great cathedral." Which man are you? - (Charles M. Schwab)

Looking at this bigger picture, we need to identify the most probable sequence that will take us there. Moreover, very importantly before it gets too crowded. This might mean revamping the organization on two major fronts: Operational and Cultural. Cultural (out of scope for this article) will follow once the operational starts rolling.

The path should be clearly defined, directed, scoped and with a time-bound objective end insight.

First, the right set of operational overhaul needs to happen in four major areas of IoTA enablement:

  1. Data capture – Whether we act as the primary hoarder or downstream, the data deluge will flood us all, hence timely shift to proper data-archiving technology will be essential.
  2. Data quality/value – With increased diversity in data sources and types, identifying the useless and misleading portions will become very important; otherwise, data value density will fast degrade.
  3. Data transportation/transformation – Integration, aggregation, and representation of the sanitized data are the staple source for a super-fast interpretation layer. The efficiency of this layer will determine that of the next.
  4. Data interpretation – With diversity comes complication as we move from a 3D matrix to a multidimensional matrix (e.g., to existing transactional, demographic, and behavioral data we are adding emotional, sentimental, habitual, and social pattern data—all of which are defining one entity: the customer).

Part 2 of the blog will continue this discussion into proposing an objective preparation road map supported with examples and use cases providing glimpses of the future IoT and IoTA will carve for us.

The content, examples, and terminologies provided in this article are for information purposes only, summarized from articles available in the open internet and internal documents. I make no representations as to accuracy, completeness, currentness, suitability, or validity of the examples cited here. All information is provided on an as-is basis, which are liable to change in this fast-evolving world.

About the author

Souvik Das

Souvik Das

Senior Manager, Marketing Solutions and Research, Genpact Analytics

Follow Souvik Das on LinkedIn