Get started Bring yourself up to speed with our introductory content.

Commercial IoT environments require shared edge infrastructure

As enterprises continue to implement IoT for customer experiences, smart building systems, automation and more, we see that the workflow from the sensor to the cloud often exists within its own, individual silo. The result of this is that many businesses today are operating multiple, siloed, IoT apps — each with its own devices, sensors, IoT gateways and connectivity protocol.

This is a major cause for concern. A siloed approach results in excessive up-front costs, a higher total cost of ownership, system integration issues and increased risk associated with trying to maintain security practices across disparate siloed systems. Simply put, this status quo is a mess. If we don’t do something to fix it, this setup could significantly hold back the commercial IoT industry.

As an industry, we need a solution that can serve commercial IoT applications more efficiently, less expensively and more securely. And because commercial IoT features potentially thousands of devices running across a geographically diverse area, we must provide technologies that are more robust than consumer-grade IoT, less costly and less complex than industrial IoT, and more scalable than both.

The solution sits in stopping the construction of a silo before it is built, in rethinking our project designs and by working from the data down, rather than from the device on up.

How a silo gets built — and how to stop it

This fix begins at the IoT project design level, where teams often start at the bottom of the IoT stack — with the devices — and work their way up to the cloud without consideration for other devices or “things” that might want to be tracked or monitored. Thus, a silo is born.

This is often the direct result of prototyping a single-use case system. A lone app is created with a specific goal and budget. Then “pilot-itis” ensues and each new app — for lighting, automation, security, tracking and so forth — is developed in its own silo, walled off from other apps and development teams.

This is not uncommon. Within commercial IoT teams, there are often several prototyping efforts taking place concurrently. Each effort has its own version of a bottom-up infrastructure in support of its unique use case. This bottom-up approach to IoT infrastructure makes coordination, information sharing and business integration difficult at best. Further, this silo mentality is an expensive and unsustainable proposition, presenting security risks and fragmented data.

It is at this design level where we must put a stop to the construction of these silos. We need to introduce a process that reduces complexity, eases integration and maximizes the scales of economy. The need to shift away from the device-up view and toward a data-down approach is imperative.

The data-down approach

Utilizing the data-down approach to developing a commercial IoT infrastructure can break these silos down — or stop them before they are constructed.

During development, treat the IoT workflow itself as a design component that is equally as important as the data. Instead of asking what devices and sensors are needed and how the data will be gathered, ask instead what kind of data types and elements are needed and why. What will be done with the data? This new direction in thinking begins in the cloud and naturally trickles down to the edge devices.

Once at the edge, the gateways can be considered. Ask what type of edge computing is needed to support the collection of that data and possibly future data requests. Gateways are the hub of commercial IoT activity, so consider the types of connectivity required for devices (downstream) and cloud integration (upstream). Make sure the gateway can handle the specific requirements, both now and in the future, for shared infrastructure and interoperability. Does the gateway:

  • Have flexible connectivity options such as the latest version of Bluetooth, Zigbee, Thread, Wi-Fi and so forth? Flexibility here is key so you are not playing catch-up with device changes down the line.
  • Use containers in a way that allows multiple apps to run simultaneously without interfering with neighboring apps? Containers also play an important role in securing the edge-app infrastructure by encrypting app storage and verifying software signatures from boot to application launch. End-to-end security is paramount in an IoT environment.
  • Monitor performance with intuitive tools and analytics? Does it orchestrate updates? Without these remote management abilities, support teams will struggle to effectively manage edge computing, connectivity, updates and security patches.
  • Allow for edge computing? When data between systems can be shared (because it is not siloed) at the edge, efficient applications can work together before the data is even pushed to the cloud. This allows for distributed workloads and faster response times using local command and control. An example would be for the room occupancy system to work with the lighting and building management system, and possibly the security system, all within the gateway as someone enters a room based on local rules without having to go to the cloud.

A practical example

How does this approach affect the end user? Consider a smart commercial building — be it retail, offices or quick-service restaurants — where there are multiple systems in place to keep the building open and running on a daily basis. Security, lighting and sensing technology are just a few of the IoT applications to be considered. When built from the ground up, there will be a silo for each of those applications. This is a management nightmare when the time comes for upgrades or patches to those applications.

By designing the IoT infrastructure from the data down, complexity and cost decrease significantly, allowing for more efficient building support. It minimizes management chaos and facilitates system integration of IoT systems. In turn, enterprises can maximize business value while reducing time to market, risk and cost.

Ultimately, this single infrastructure framework will allow enterprises to install, run, manage and update multiple edge applications as needed. It will reduce the complexity, cost and risk associated with these actions as well.

The correct solution

Typically, it is best to seek out a manufacturing partner that treats the infrastructure as its core competency. A firm designing an IoT application should focus its time on its expertise — that is, the application itself. Meanwhile, a firm focusing on hardware and infrastructure is uniquely positioned to handle the complexity, risk and interoperability these networks demand.

Baking an IoT application into the right hardware and partnering with the right solutions provider will not only solve the silo issue for your business, but will also allow for rapid and streamlined deployment of your system.

When seeking out such a partner, ensure its product addresses the previously mentioned requirements. It must be interoperable in terms of connectivity, it must prioritize security both in its system architecture and its long-term support, and it must allow for monitoring of performance and orchestration of updates.

Designing IoT applications the right way, and utilizing the right hardware, is key to avoiding the trouble with silos. The future of commercial IoT demands we all take interoperability and security seriously while also allowing for a shared infrastructure that can run multiple edge applications. The technology exists to achieve this outcome. It’s up to us to build the appropriate ecosystems with that technology and truly empower commercial IoT.

All IoT Agenda network contributors are responsible for the content and accuracy of their posts. Opinions are of the writers and do not necessarily convey the thoughts of IoT Agenda.

Data Center
Data Management