- March 16, 2017
Ready for Action: Moving Beyond Data NoiseShare this:
By Marc Wilkinson, Chief Technologist and Mobility Global Practice lead, Enterprise Services
This is the second in a series of articles and insight from Marc Wilkinson on IoT.
Enterprise data is like a stadium of people shouting. It’s hard to pick out the relevant voices from the noise.
Organizations are being overwhelmed by a wealth of data. It streams in from sensors, social media, transaction processes, and customer interactions. To move beyond the resulting noise, enterprises must develop strategies to tease out the intelligence locked within. It starts with formulating a solid business objective and then understanding the data required to reach that objective.
What’s Your Problem?
Consider a hotel that has determined it can generate value by accelerating and streamlining the guest check-in processes. To figure out what data it needs, the hotel could run experiments to determine how front desk staffing levels impact the process. It might measure the effects that variations in lobby temperature, lighting, and music have on check-in times. It might also consider testing a mobile app or a rewards program card with an RFID chip that identifies guests and displays their information on a screen before they enter the lobby.
By clearly defining the problem you want to solve and what you want the data to accomplish, you reduce noise while boosting the visibility of actionable information.
The Power of IoT
The interconnected network known as the Internet of Things (IoT) is fueling new product and service development by generating rich data sets embedded with valuable insights. IoT can enable new incremental revenue streams through the development of smart products and services, for example. IoT can also optimize existing operations by delivering near real-time insight, enabling enterprises to automate and operate in an on-demand fashion.
And because IoT is all about connecting data sources to business applications and processes, it can drive cost efficiencies and productivity improvements. Areas realizing the biggest gains include capex, labor, and energy. But the real value of IoT lies in its predictive analysis. This is clearly visible in processes such as car and aircraft engine diagnostics, where predictive capabilities reduce costly downtime and extend the operating lifetime of capital equipment.
Yet IoT is not just about enterprise operations and business improvement. It’s also about making our personal lives easier. Through devices such as wearables, we can anticipate and address health issues before they result in a costly trip to a specialist. Ready access to information allows people to plan and act quicker with better results. This holds true whether one is grabbing an umbrella 30 minutes before a storm strikes or boosting widget production by 0.0017 percent to match demand signals.
Sharpening Data Points
Reaping rich benefits from IoT demands a deep understanding of the data you’re collecting. Data has a life cycle, a period of currency that needs to be baked into the way we collect, store, and use it. Data scientists must understand the data source, its accuracy and relevance, and the level of granularity required to achieve the desired outcome.
The oil and gas industry, for example, sought greater insight by deploying sensors with a higher degree of sensitivity and accuracy. But more accurate sensors meant more robust data streams. They soon discovered that they were receiving far more data than they could realistically analyze and process—defeating the purpose of more powerful sensors.
Today, we’re able to build analytics into all kinds of systems because it’s relatively easy to do so. But these analytics are all but useless without a solid understanding of the problem the data is tasked to solve and the level of accuracy required. Automating decisions or harvesting insights from data that isn’t aligned with its business use leads to poor decisions and worthless insights. That’s why it’s crucial to start with a solid understanding of the business challenge you’re attempting to overcome and the nature of the data you need to fuel solutions.