Evaluate Weigh the pros and cons of technologies, products and projects you are considering.

Key factors to consider in building your IoT network

According to the most recent Ericsson Mobility Report (November 2018), the number of cellular-based IoT connections is likely to grow from 2018, by 27% annually, to reach 4.1 billion by 2024. The challenge for CIOs, CTOs and their operational teams is how to access a connectivity infrastructure and ecosystem that can support all these devices.

When figuring out the best approach to take in building an IoT connectivity strategy, there isn’t necessarily a perfect single strategy — it really depends on the use case and the nature of the devices to be connected. To help break it down, consider the three W’s — what, where and why. What is the device to be connected?; why do I want to connect it? (in other words what data do I need to extract); and where will these devices be deployed?

Let’s examine these questions and some other key elements, a little closer.

Device tracking and data extraction

You’ll need to understand what type of data is being collected and how often the device is communicating back to a monitoring or analytics application, which will increasingly be in the cloud. If it’s a large amount of data that is needed in near-real-time, then edge computing may be required to process information more quickly than sending it to the cloud.

Next, consider the location of the devices, both from a geographical and physical standpoint. For example, are the devices going to be deployed in a specific site — like a factory or campus – and will they be in the same country or possibly be deployed in multiple countries? Also think about the device or asset itself. Does the thing to be connected have access to a power source (like a car via the battery or a vending machine that’s plugged in) or does it require its own power/battery pack? Is the device easy to access and replace the battery, or will it be in a remote location, where battery life becomes critical?

The location and type of data that is being passed back and forth also determines the type and level of security that is needed. Use of near-field communication (NFC) or some types of low-power wide area network (LPWAN) can mean that security is compromised. Cellular connectivity brings a legacy of secure device access, identity management and authentication techniques, combined with security techniques that have been developed to protect IP-based communication services. This creates a secure foundation for end-to-end IoT communications. Of course, cellular connectivity is not always appropriate for IoT deployments. A fitness tracker that is connected with an app on your smartphone only needs an NFC connection, such as Bluetooth, to become connected, while a car or truck may have a range of connectivity requirements that define the need for a pervasive, cellular connection. It is important to identify the best type of network access for the IoT system, whether cellular, satellite, Wi-Fi, private LTE, Bluetooth, LoRa or some other type of LPWAN connection. It is clear that the level of security, coverage requirements, communication costs and impact to the hardware and battery design should all have equal consideration when designing an IoT connectivity strategy.

Relationship to the network

Another important factor is the interdependency between the device and the network. The more basic a device is in terms of the data being recorded and the data transmission rate, the more likely it is not to be fully secure and potentially not to work across multiple network types.

On the other hand, the more flexibility and capabilities that are built into the device (or the device-side application), the more capabilities can be meshed between the device, application and network. But this typically requires more memory/OS on the device and higher data transport over the network — which means higher battery usage.

Network technology and connectivity decisions are very important factors to consider as an integral part of the overall strategy for any IoT project. Will the devices rely on one type of connectivity, or will they need to be able to interact with multiple connectivity options?

Partner, partner, partner

Because there are multiple elements to delivering an IoT system, it is imperative that an enterprise either works with a systems integrator to ensure the various components work together or selects providers with a clear reference model with other partners and have experience in cooperative work.

At Tata Communications, we are partnering with semiconductor companies, chipset manufacturers and other OEMs to ensure our platform works with a variety of devices and applications. We are working with vertical solution SaaS providers and public cloud service providers (including Microsoft Azure and AWS) to ensure secure connectivity all the way from the device to the cloud and back again. An example is the work we are doing with Thailand-based fleet management specialist DRVR, helping transportation companies in Asia better manage their vehicle fleet activity and save on fuel costs. DRVR’s connected sensors can identify location, driver behavior, engine idling time and other parameters to alert drivers or fleet operators to take action.

Expect the unexpected

It is in the operational details where connectivity strategy and network selection are critical in helping to build a successful IoT project. However, there are many considerations, such as unexpected delays in testing and deploying a system are common. In what remains a very fragmented market, it is vital to ensure end-to-end interoperability of the various components — device, network, cloud, application, analytics — work together, to ensure that an IoT project can be successful from proof of concept all the way to growing at scale.

Large-scale, multinational IoT projects are still relatively new and immature. As a result, nobody should assume that connectivity “just works.” The key is to have a technically complete plan, consider the key factors and, perhaps most importantly, be patient and work through the problems with qualified and like-minded partners in order to optimize the business case or the return on investment.

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