Simplifying the Internet of Things Conversation
The IoT is one of a number of new sources, along with social media and wearable computing, which can be combined with data science, collectively as the Big Data Killer App for organizations.
Step 2: Identify The Decisions. The original chart states that “Layer 7” is about “transformational decision making.” However, instead of making this layer 7, I’d make it layer 2 (or step 2). You need to identify the decisions the key stakeholders need to make in support of the targeted operational initiative (see Figure 3).
By the way, decisions do not need to be transformational to deliver compelling business value. Optimizing boring decisions such as when a vehicle or jet engine needs servicing or how best to load balance the network given a spike in demand can also deliver a pretty nice ROI.
Step 3: Identify Technology And Organizational Requirements. With the targeted operational initiative and supporting decisions in hand, the rest of the IoT technology and organizational requirements quickly fall into place, including:
- Who are the stakeholders (technicians, mechanics, engineers, logistics managers, supply chain managers) impacted by the targeted operational initiative?
- What descriptive, predictive and prescriptive questions do the stakeholders need to answer in support of their key decisions?
- What additional data sources – both internal as well as external to the organization – should we consider (e.g., weather, traffic, technician notes, product specifications, field problem reports)?
- What actionable recommendations (to the decisions that they are trying to make) do you need to deliver to the stakeholders in what timeframe and in what manner?
- What data architecture and technologies do I need to support this process?
By the way, I recently covered this process in more detail at my “Developing A Big Data Business Strategy” session at Strata + Hadoop World on March 30 (see Figure 4).
Step 4: Contemplate “Right Time” As Well As “Real Time”. Not all decisions need to be made in real-time. For example, the article states:
…a company managing a fleet of delivery trucks can detect when truck parts are performing suboptimally and schedule these trucks for preventative maintenance long before the vehicle breaks down. These types of real-time decisions greatly reduce maintenance costs and improve the overall delivery performance of the fleet.
Determining and scheduling preventative maintenance is not a real-time decision. I am not going to rush a team of mechanics and parts to wherever the truck is located at the first sign that the vehicle is going to need maintenance. If I can predict that a truck is likely in need some maintenance, then I probably have a good estimate as to how soon I need to schedule that maintenance. And then I can schedule multiple maintenance activities to better reduce my maintenance costs and downtime for that truck (or wind turbine or jet engine or power generator or etc.).
The “Big Data Killer App” is the ability for organizations to couple new sources of data, such as social media, wearable computing and IoT, with data science to make better decisions. When you start aggregating all of those decisions across multiple use cases, then you have something that could be truly transformational to the business (see Figure 5).
 A “killer app” is any computer application that is so indispensable that it proves the core value of some larger technology. In this case, the IoT is being positioned as being that application that is so indispensable that everyone must adopt big data for their very survival.
Bio: William (Bill) Schmarzo, the "Dean of Big Data," is the CTO of EIM Service Line at EMC. An avid blogger, Bill speaks frequently on the use and application of big data and advanced analytics to drive an organization’s key business initiatives.
Original. Reposted with permission.
- How IoT is Jeopardizing Your Business Security
- Deep Learning for Internet of Things Using H2O
- Whether Evolution or Revolution, The Internet of Things is Here to Stay