IoT deployment is expanding out from consumer-based applications such as smart home devices and wearables to applications in the areas of public safety, emergency response, industrial automation, autonomous vehicles, and the Internet of Medical Things (IoMT). Many of these are mission-critical processes with life or death implications — so it’s essential to get it right.
In this article, we’ll be considering some strategies and best practices for IoT deployment, and looking at the challenges that must be overcome before, during, and after implementation.
A Three-Part Strategy for IoT Deployment
Several factors come into the successful deployment of an IoT system and its connected devices, including security, interoperability, power or processing capabilities, scalability, and availability. A general strategy for implementation consists of three phases:
1. Consult
Having determined whether you will buy your IoT infrastructure or build it out yourself, it’s essential to first consult with the internal or external team responsible for running the project and establish a road map for your IoT deployment.
You’ll also need to consider factors like your current state of readiness and what the project needs to return to justify its cost. Your “readiness assessment” should include calculations of cost, value, and ROI (Return On Investment).
At this stage, it may be a good idea to conduct a pilot program testing the validity of your ideas and providing a foundation for the strong business case you’ll need to present to management when negotiating for the required project resources.
2. Develop
During the development phase, you should optimize the initial plans created during the consultation, for performance, cost, and quickest route to market.
This is also the period during which you or your external contractors will be building out the cloud platform and setting up the analytics framework. These processes will typically involve configuring analytics, cognitive services, artificial intelligence (AI), and machine learning (ML) platforms to tackle features relevant to your business, and developing the interfaces for consumer-facing web platforms, applications, and data visualizations.
Consideration should also be given to device management, connectivity, and how these features will change during the life cycle of your IoT deployment.
3. Deploy
The consultation phase should have yielded assurances that your development team or external contractors can install, support, and service all the necessary infrastructure and equipment during the crucial period of initial implementation.
The deployment phase marks the culmination of all your plans and provisions for security and data privacy, testing and enablement of interoperability, and forward stocking for backup hardware, customer support, technical support for Wi-Fi or cellular networks, gateways, web interfaces, apps, and your cloud platform.
IoT Deployment At Scale
With smart cities, connected global enterprises, smart transport systems, and ecosystems of wearable devices on the horizon, attention has also to be given to large scale IoT deployments. Key factors to consider relate to five main areas:
1. Devices
For larger-scale IoT deployments, expect a mix of new and legacy devices employing different technologies and serving multiple purposes. This variation makes interoperability a key requirement.
Any aspect of the deployment that works exclusively with only one vendor, one platform, or one technology should be avoided. This will safeguard against incompatibility problems and proprietary lock-in.
2. Connectivity
All devices on the network need to support your preferred connectivity standard. While many deployments default to Wi-Fi, a project may have specific needs or require connectivity in areas where Wi-Fi isn’t available. In such cases, cellular connectivity may be a preferred option, as it typically offers low power consumption and long battery life for devices, and greater accessibility.
3. Device Management
With an IoT deployment involving thousands or even millions of connected devices, performing even basic operations like configuration, security patches, and maintenance may be problematic — especially if physical access to devices is limited.
Connectivity, industry standards, and protocols assume vital importance in device management for large-scale deployments. Regulated open standards developed with the IoT industry in mind are the safest bet. Strong, flexible, and reliable protocols will enable your device management solution to cope with the sheer scope of the deployment and any changes that may affect it in the future.
4. Data Processing
Though it’s tempting to parcel data from a large-scale deployment into manageable smaller packages, this can lead to the creation of silos with no connection to each other.
Instead, all data should be aggregated and accessible via a centralized platform. This will also give system administrators a “big picture” view of the deployment.
5. A Flexible Management Platform
A myriad of devices with multiple sensors and actuators collecting reams of data every second of every day requires a versatile platform to manage it all. This platform should be flexible enough to accommodate different solutions and able to adapt to future changes.
IoT Deployment Challenges
The principal IoT deployment challenges are often summarized as the “5C’s of IoT”, as follows:
Connectivity
A seamless flow of information to and from devices, infrastructure, applications, and the cloud is vital to successful IoT deployment — particularly where mission-critical operations are involved. The complexity of wireless connectivity with its still-evolving set of standards can complicate matters as much as managing a diverse set of devices.
Flexible design and testing solutions capable of assessing devices with many radio formats are necessary for meeting this challenge. Solutions should be simple, low-cost, and capable of application during both R&D and manufacturing phases.
Continuity
This aspect mainly concerns battery life and methods of extending it for IoT devices. Long battery life is a desirable attribute for consumer IoT goods. Battery life of five or ten years is commonly expected for industrial IoT devices. And for medical IoT devices such as pacemakers, battery failure simply isn’t an option.
To prolong battery life and ensure continuity, integrated circuit (IC) designers need to design the circuits with deep sleep modes that consume minimal current and reduce clock speed and instruction sets. Systems that operate on low battery voltages are also desirable.
IoT Deployment Challenge: Compliance
There are radio standards and global regulatory requirements to which IoT devices must comply. Compliance testing includes radio standards conformance and carrier acceptance tests, and regulatory compliance tests such as RF, EMC, and SAR tests.
Testing can be complicated and time-consuming, often requiring designers and manufacturers to seek in-house pre-compliance test solutions to meet product release schedules.
Coexistence
Wireless congestion or the overcrowding of radio channels is a logical consequence of billions of IoT devices competing for bandwidth. Standards authorities have developed testing methodologies to evaluate device operations in the presence of other signals.
For IoT deployment, coexistence testing is crucial in measuring and assessing how a device will operate in a crowded, mixed-signal environment and assess the potential risk to maintaining wireless performance in the presence of unintended signals.
IoT Deployment Challenge: Cybersecurity
To date, little has been done to address the over-the-air or OTA vulnerabilities afflicting IoT devices. OTA and potential endpoint vulnerabilities should be identified using a regularly updated database of known threats or attacks, and devices should be tested on the basis of these same criteria for response and anomaly detection.
