We sometimes think of the Internet of Things (IoT) as a monolithic solution that companies are rapidly deploying to collect and analyze sensor data in order to do amazing things.

The reality is that today’s IoT vendor landscape encompasses a bewildering assortment of devices, gateways, storage systems, analytics solutions and SaaS services, along with a myriad of protocols, security standards and data models. Because of this, matching customer requirements to the right solution set takes time and research, especially when it comes to three critical areas where many vendors and customers have failed to anticipate increasingly common, high-priority use cases.

The Last Mile

Current IoT solutions are relatively efficient at collecting sensor data in gateways, moving the data to storage solutions, performing predictive analytics on the huge data stores, and communicating back to the gateway if actions need to be taken. However, these solutions typically don’t communicate with other vital systems customers rely on to run their businesses, including customer relationship management (CRM), enterprise resource management (ERP), human capital management (HCM), patient relationship management (PRM) and more.

The customers I talk to insist that such communication is essential if they are to create end-to-end business processes that result in people taking actions based on IoT-driven data collection and analysis.

For example, many devices from different vendors currently monitor patients in their homes to detect declining or adverse conditions. These devices, which monitor for weight, temperature, blood pressure, blood sugar, mobility and more, likely send data to multiple vendor healthcare clouds.

To use this environment to accurately determine if a nurse practitioner or doctor needs to be dispatched to the home, there needs to be a single platform that can consume all the data and alerts from these disparate devices and interface with the appropriate Patient Relationship Management (PRM) system that can initiate the required response.

Hybrid IoT

Many IoT vendors create vertical solutions designed to quickly show value. In a matter of minutes or hours, they can deploy sensors, set up gateways, send data to the vendor’s cloud storage, perform analytics, and deliver a dashboard using pre-defined rules. But what if the analytics application that is part of the vertical solution fails to meet a customer’s needs? What if the customer already has a preferred analytics application that is integrated with other business processes?

In these cases, getting the data collected and stored by the IoT solution into the customer’s existing analytics application may require a data scientist to transform and normalize the IoT data – usually requiring a lengthy and costly IT project. Similarly, vendors such as Salesforce, Amazon and Microsoft are bringing their own cloud-hosted IoT solutions to market, allowing customers to send device data to them and obtain quick access to predefined dashboards.

But what if the customer wants to upload additional data, such as customer or product information, to the cloud-hosted IoT solution for a more robust analysis? And what if a company can’t afford the time or cost of a data scientist? IoT solutions must be flexible enough to meet the real-world data needs of a growing and evolving business.

IoT Analytics at the Edge

As noted above, current IoT solutions typically send the collected data to a storage solution for the purpose of predictive analytics performed on very large data sets. But customers I talk to see the potential for a very different use case: quickly analyzing a much smaller set of data to achieve a more immediate response.

Consider a cold storage warehouse that wants to continually analyze just the last few minutes of data from a group of sensors in the entire warehouse, or maybe just one zone, in order to detect and react to any recent changing conditions. This could be for sudden temperature or humidity changes, equipment anomalies such as failing sensors, or even a security breach due to malicious behavior. Shouldn’t there be a way to perform this limited analytics on the IoT edge gateway itself?

IoT edge analytics would enable far more reactive scenarios compared to big data analytics, which is more suitable for proactive and predictive scenarios.

Key Considerations

Here are the five top considerations for organizations that want to ensure their IoT initiatives will meet their evolving business goals.

1. Think long-term about possible use cases and what other data and systems might be involved.

2. Evaluate vendor solutions for data format compatibility, how data is consumed and analyzed, and how results are communicated. If you are using a cloud-hosted IoT solution, make sure data integration can take place before data is uploaded. Also consider whether the vendor’s data governance policies are aligned with your organization’s policies.

3. Assess whether vendor IoT gateways have the flexibility to support edge analytics.

4. Leverage third-party cloud technologies, such as integration Platform as a Service (iPaaS), to integrate IoT solutions with existing business processes.

5. Identify the location of all data sources, both business data and IoT device-generated data. This information is essential for optimizing reactive, predictive, and proactive scenarios, as well as for increasing the value of data integration.

(About the Author: Michael Morton (@michaeljmorton) is the chief technology officer at Dell Boomi, where he drives product direction and innovation. He has been leading and producing a wide range of enterprise IT solutions for over 25 years.)

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access