IoT Strategy and Planning

Cloud-enabled IoT implementation is one of the key digital transformation initiatives considered by many organizations. Reducing costs, growing revenue, enhancing operational efficiency, and improving safety are some of the key goals for organizations to initiate IoT-based digital transformation initiatives. The fundamentals of your strategy will involve understanding your requirements, running a pilot, deciding on out-sourcing, choosing the right solution, and considering edge computing.

47% of organizations will increase IoT investments despite the COVID-19 pandemic according to a Gartner report.

Understanding the requirements

An organization’s vision will drive the IoT strategy and will also help in landing the requirements for the IoT initiative. There are some basic questions and considerations that need to be addressed upfront so the right IoT system can be built which will fulfill the organization's needs.

• Is there already an existing IoT network?

- Are the devices compatible with the selected cloud provider?

- How good is the security of all the components?

• Is there a need to generate real-time actions from your IoT system?
• Is there a requirement to store all the IoT devices generated data for future analysis and insights?
• Does the planned IoT system require customized development?
• Is there in-house expertise to set up and maintain the IoT system?
• Is there a requirement to create a digital twins system for the physical IoT system?

Once these basic questions have been answered, they will help in building the right strategy for your IoT system.

Pilot to production

Assuming there are internal IT resources available or even if it has been decided to out-source the IoT implementation to an external vendor, it is recommended to do a pilot for the eventual production system that needs to be built. Doing such a pilot will not only help you understand the end-to-end system but will expose some additional considerations and requirements that you may have to consider as part of your overall strategy and planning.

During the pilot phase, the cloud IoT services can be configured with a set of experimental devices. Not all current IoT devices or the new devices which need to be connected to the IoT infrastructure may be compatible with the selected cloud provider. Each cloud provider keeps a list of compatible devices, which have been certified for their cloud, and this compatibility should be checked upfront.?

Security is always a key area of consideration for an IoT system, and all aspects of security should be considered and understood in the pilot phase. This will include devices, IoT infrastructure, edge network, on-premises network, cloud infrastructure, and IoT cloud services.

In-house or out-source

Depending on the complexity of your IoT system and the availability of IT resources, a decision needs to be made whether to build the IoT system utilizing internal resources or to out-source the initiative to a third-party vendor. IoT systems are generally very complex in nature and have many components. While your IT staff may understand some aspects of the IoT system, they may not understand all the components or how the system functions end-to-end. For this reason, it is recommended to work with a certified vendor with experience in multiple IoT implementations. Assuming you are going with a cloud provider to use their IoT services, the cloud providers keep a list of partners who can help other customers with IoT implementations. One of these partners can be considered for your IoT implementation. If there is already a vendor of choice, who is maintaining your current IoT system, you want to ensure that they are fully certified to do cloud implementations.

One practical approach to consider would be to bring in a partner for initial assessment and deployment along with retaining them in the long term for helping with complex maintenance issues. While you are going through the assessment and deployment of the system, train your internal IT staff to understand the system so they can manage and maintain it in the long run.

Type of solution

Before a solution is chosen from a cloud provider, there are some key questions that need to be answered upfront to help pick the right solution offered by the cloud provider.

Basic infrastructure decisions that should be made upfront include:

• Are you utilizing your current IoT vendor’s infrastructure for managing your existing IoT network?
• Do you want to set up your own IoT network and connect it directly to your cloud provider?
• Do you want to connect your own IoT network to your on-premises infrastructure in the middle and then connect to the cloud provider?

Cloud providers offer various types of IoT solutions for their customers. Microsoft Azure Cloud as an example provides Azure IoT Central which is a SaaS solution and can be used for low to mid complexity IoT systems.?Microsoft Azure also offers a wide range of IoT services that can be combined to build a fully customized solution. Azure IoT Central can be a great option for organizations that do not have developers or limited developer resources to work on the IoT initiative and want to get started quickly on a basic IoT system. While easy to set up and configure without developer experience, this solution, however, does not have the flexibility to be customized for all your requirements.??

Expansion and scalability

If you are going to expand your business globally and with that, your IoT network in those regions, one of the first considerations should be if the cloud provider can support you with IoT services in the new global region. If there is a plan to start small with some IoT devices in one region but to grow big in the future, make sure that the cloud provider can offer the required scale in their IoT services to support the future expansion of IoT devices and infrastructure.

One of the key features of cloud computing is that it offers both scale and global expansion capability as the organization’s requirements grow. While a small network of IoT devices can be supported and managed via an on-premises or a third-party vendor network using a proprietary IoT solution, growing this IoT system may not be possible. Cloud IoT services can help you to scale the system and to support hundreds or thousands of devices, expand the IoT network globally, or create a new global IoT infrastructure. With IoT services available in all the key regions of the world, cloud providers can come to the rescue to sustain your global IoT expansion.

Enabling edge computing

If an IoT use case is going to include temporarily disconnected IoT devices along with a requirement to take real-time business-critical actions, then an edge computing network should be set up.

Edge computing is a relatively newer development in the IoT world and typically applies to already existing mature or advanced IoT implementations with a need for real-time actions. Enabling edge computing will significantly reduce or eliminate the latency in sending all the IoT data back to the cloud for processing and analysis. See Figure.

Edge computing allows for close-by processing of data on the edge devices which leads to real-time actions getting triggered from the system. Edge devices typically have ML models in them which are used to process IoT data and trigger actions. For temporarily disconnected devices, edge devices offer storage and the ability to upload to the cloud upon network connectivity restoration.