From IoT architecture to IoT deployment: Key elements for success

The IoT ecosystem is highly complex and although adoption rates are soaring, many organizations still struggle to build, deploy and maintain IoT systems that deliver tangible business results. In some cases, companies become excited about the potential of IoT, but do not align their projects with well-defined business outcomes, such as increased operational efficiencies or new revenue-generating service offerings. The result: stranded investments in IoT projects that did not produce bottom-line results.

To avoid wasted funds and negative sentiment around IoT within your organization, the following considerations are critical:

• What are the business objectives?
• How will we design the IoT system to meet these objectives?
• What are the global ramifications?
• How do we ensure security?

It is important to determine your business objectives at the outset and define how IoT will be used in your organization to align with these objectives. The next step is to conduct a transparent and comprehensive assessment of your organizational IoT readiness, providing the contextual information needed to generate an attainable IoT strategy. From there, you need to construct an IoT architecture that aligns with your IoT strategy, specific technology components and milestones to business objectives.

Designing your IoT architecture

IoT solution architectures are more complex than most traditional IT projects because of the many layers of complexity at the device, network connectivity and application levels, in addition to the interoperability among each. Architectures can be viewed from a logical OSI stack perspective or an architectural end-to-end perspective. Both are relevant when assessing your requirements.

From a stack perspective, an IoT solution is like a three-layer cake. At the bottom is the connectivity, for which there is a plethora of options, depending on the business use case. The next layer up is comprised of two components: device management and the management platform. The devices and endpoints connect into the network and the platform manages the devices and processes the data. The top layer includes the applications and data visualization, which deliver the insights that result in business value.

From an architectural end-to-end perspective, an IoT system begins with an endpoint device, such as a sensor, that collects data. That data is transmitted via network connectivity, such as cellular or Wi-Fi. These components are managed by a platform, or multiple platforms, that enable connectivity management, device management, application development, security and other functions. The data is then stored in either a cloud or hosted data center, where it can produce analytics or feed back-office systems.

Tackling global considerations

Global deployments bring additional complexities. Navigating IoT device certifications and sourcing connectivity options across multiple regional networks and carriers can be complicated. For example, a global system might need to work on a 2G network in Europe, a 3G network in LATAM and an LTE network in North America. If the device were transported across any one of those networks, it would need to work seamlessly without any modifications to the device itself.

It is important to understand global nuances and how they translate to costs. In the example above, your organization could deploy one — potentially expensive — device that provides seamless transition capabilities, or utilize multiple devices, each specific to a given region. In most cases, the latter option is significantly more cost-effective.

Overcoming security challenges

Securely managing IoT systems as they scale requires forethought. Common challenges include sensor failures, bot attacks and camera hacks. Enterprises that deploy IoT devices into their customers’ premises are responsible for ensuring that the device does not become a vulnerability for the end user. This end-to-end security is a key challenge for the industry overall, but especially for large-scale deployments.

Ensure security measures are included in your IoT system by performing threat modeling during design. Threat modeling begins with an architecture diagram of the system that clearly depicts how data flows throughout the application’s different elements and where attackers can break in. Then, compile a list of all threats associated with the solution, such as the aforementioned bot attacks or camera hacks. After the threats are identified, prioritize and mitigate each at the design stage, before the IoT system is introduced into the field.

Getting started

It is important to build a results-oriented IoT strategy from the outset. Determine business objectives first, and then choose the right technology for the processes and goals to achieve your desired results and maximize returns on IoT investments. Your IoT strategy should be reflective of your business strategy, and each individual IoT sub-strategy should align IoT technologies to targeted business processes. A focused approach allows your organization to identify, on a global basis, where the processes operate, the back-end systems that need to be integrated and the levels of security required.

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.