How (NOT) to Fail With IoT and AI


The majority of the growing industrial companies today are betting their biggest bets on tackling diverse technology initiatives in the light of Industry 4.0 transformation. 

The IoT universe is manifesting newer forces that are erupting at breakneck speed. 

On the one end, there is an explosion of smart sensors, coupled with a Tsunami of big data. On the other end, there are numerous, complex big data platforms to collect, process, and store data at a mass scale with groundbreaking advancements in technologies. This further presents a deeper ‘all-too-human’ problem where our ability to store data scales while the ability to process and analyze it doesn’t

The Industry 4.0 era requires specific operational methods and technologies that allow us to create new value. This value is derived by tapping the data from trillions of connected machinery and speeding up the path to gain access to that insight.

However, the world today is in the face of significant data deluge problems to scout for valuable insights amidst unorganized and siloed data mines. But thanks to IoT and AI, the future points towards a paradigm shift where industries can experience the benefits of greater asset availability, better sustainability, and higher production efficiency.

Stepping Up the IoT Game to Break Conventional Industrial Norms

Leveraging data from IoT devices help steer and accelerate the path to end-to-end industrial digital transformation. Industry 4.0 aspirations are far from on the ground industrial digital and data maturity. Most organizations struggle to get the right data at the right time to generate valuable insights to facilitate faster, accurate business decision making. 

Here’s why the status quo doesn’t work for Industry 4.0 and how companies can rethink their data strategy for achieving rapid and seamless production rollouts:

Absence of Enterprise-Wide IoT and Data Strategy

With the Industry 4.0 change, companies are heavily investing in numerous data science and machine learning initiatives to come out as Industry 4.0 champions. They are looking at finding something that may or may not even exist. 

According to Microsoft’s recent analysis, over 30% of industry leaders are stuck in this vicious circle of aimlessly repeating one IoT initiative after another. This gap arises when companies fail to see that holistic enterprise-wide industrial digital processes require data strategy, data maturity, and strategic business KPIs.

Businesses today thrive not on cursory and tactical data, which reveals if there was a profit or a loss but instead on much more in-depth insights. For example, by what percentage has the profit margin shifted, what are the influencing factors, and how exactly has this impacted bottom line. 

The truth is that making this quantum leap from pilots to actual production is a mammoth effort! Still, yet most of their PoCs don’t translate into production as they do not see a significant positive bottom-line impact. 

Companies should have pragmatic models that are designed to propel enterprise-wide IoT initiatives to have valid proof-points from early successes. 

A robust data strategy starts not only at selecting the best-of-the-breed tools and technologies to tie the loose ends and stitch in the data sources together. It should rather start with the critical business priorities, key KPIs around which decisions for – what kind of data should be collected, measured, and analyzed have to be taken. 

In this 60-second video, Big Data and AI Industry Expert and my favorite, Dean of Big Data, Bill Schmarzo beautifully highlights how Big Data is not about technology but rather about the economic value of data. 

As an industrial company, which wants to ensure highly efficient factory operations, you need to think about tracking critical operational KPIs. After all, it is the sum of the KPIs put together that narrate a story about the performance of the line, plant, and organization.


 An operations KPI or metric is a unit of measurement that can quantify and optimize your operations processes based on varied volumes, time frequencies, and costs categories. 


Accurate tracking of these KPIs can provide the much-needed business and operational insights to meet your operational goals.


Anyone from the plant manager to the plant head or even the factory CFO applies these KPIs in various forms to create value across the business life cycle.

When businesses approach Industry 4.0 pilots without having a clear understanding of the KPIs and data, they are bound to end up in four scenarios listed below.

  • Impact: Unclear business objectives which mean they can’t justify what are the potential benefits
  • Budget Overruns: Lack of clear business outcomes which means budgets will overrun and they will be hard to quantify
  • Data Quality: 75% of the companies realize during the pilots about the lack of right data for solving any/specific business problems
  • ROI: Difficult to put a business case for scaling further innovation without a well-defined strategy which ranges across multiple parameters – 
    • The need to buy more tools or technologies
    • In-house data skills vs. cost of working with consulting or outsourcing companies 
    • Leadership strategy and consensus for Industry 4.0 commitment 

Better Together: Combining the Prowess of IT and OT 

Companies need to eliminate the business blindspots to put digitization into place. When companies bring the IT and OT experts in a single room, together they can bring to the forefront, not just challenges but also potential solutions for strategic areas such as automation and data strategy. This approach fosters a simplified digital culture to experiment with IoT and AI to build, learn, and use enhanced ways of operating. 

Data Scientists and IT Teams can use the treasure of OT business expertise in machines, types of equipment to help drive critical decisions and increase revenue margins to an exponential extent by drawing upon an extensive collection of statistical techniques

For example:

  • What are the trends in the data? 
  • How is a machine performing over time? 
  • What could be the leading influencers of production delays? 
  • What are the clusters or groups in the data that are driving low production volumes, and what are the correlations?

The result of an OT and IT-driven strategy is a win-win. Companies achieve higher output with lesser costs, business-relevant automation of industrial operations, and make the much-hyped move from pilots to production environments.

The Ultimate Speed Test — Achieving Agility in the Industry 4.0 Era 

AI and IoT are redefining the game for plant managers by making operational excellence a reality while simultaneously reducing costs. AI-driven advanced analytics approaches are transforming the way companies do many things – from tracking operational defects to maintaining uptime to streamlining operations. 

AI is the gateway to creating this difference for your business by harnessing the true power from massive streams of IoT data. The crux here really is not just about mere agility but converting this speed into valuable decision making and hence reducing the time to market. 

Companies must use data from IoT and AI to translate the real value of agility into decision making, cost savings, and scalable innovation with greater precision of results and faster time-to-market.

This UrIoTNews article is syndicated fromDzone