Here are the key considerations for a successful Internet of Things deployment.
Whether you want to streamline your supply chain, better manage your workforce or improve customer relationships, the Internet of Things is expected to transform how businesses operate. The idea is that with all things connected and capturing data, we can analyze and map it to business processes. The potential of IoT is tremendous, but the danger rests in poor implementation. If you don’t properly plan and execute a sound IoT strategy, you risk your competition leaving you in the dust.
As is true for virtually every IT project, the first place to start your IoT initiative is the network. Your first consideration revolves around how you will provide connectivity and sufficient bandwidth to any number of IoT data collectors. Depending on your current and future ambitions, this could mean looking at your LAN, WAN and Internet capabilities and upgrade options.
Once you understand and look past how IoT endpoints collect data in a distributed manner, what you’re left with an exercise in big data. You’re going to be collecting a huge amount of information and using intelligent software to organize and analyze this information so meaningful and accurate decisions can be made. That means that the data center will play a key role in any successful IoT architecture.
Finally, a successful IoT initiative will be one of scalability, security and sufficient support. The underlying IoT infrastructure must be able to expand for future IoT projects, data must be protected from end-to-end, and all IoT endpoints need constant maintenance.
Continue on to see how you should go about planning for your upcoming IoT initiative.
Wired and wireless LAN
A great place to start when assessing your existing infrastructure is to look at your wired and wireless LAN. In many cases, the primary concern will be to make sure you have the Ethernet port densities for wired IoT devices and sufficient coverage and capacity from a WiFi perspective. The other thing to assess is the uplinks that connect IoT endpoints to the data center where all data is collected and analyzed. Even though the amount of data collected might be small on a per IoT endpoint perspective, when hundreds or thousands of devices are transmitting data to aggregation uplinks, it can make a significant impact on link utilization. So make sure you have sufficient bandwidth today and an easy upgrade path for the future.
Depending on your IoT architecture, WAN and Internet connections may also have to be assessed. If IoT endpoints extend out to remote sites outside of the LAN, you need to make sure that bandwidth and latency is sufficient for your current IoT ambitions. If needed, you may also need to provide additional redundancy in addition to bandwidth upgrades.
Data collection methods
When planning an IoT project, it's important to understand how IoT endpoints will collect information. There are numerous methods for data collection including RFID, barcodes, Bluetooth, GPS and various other environmental sensors. Make sure you understand how each collection method works, its limitations and if the collection method is ideal for your specific implementation.
Once you've figured out your communications network and IoT endpoints, the next consideration is data storage. Here, not only do you have to think about data storage capacity, but also what availability, speed requirements and deployment model you will need. The ideal solution will be a highly scalable storage solution that handles a wide range of data types, and can handle data collection from a large number of collectors.
The question regarding analytics is highly subjective depending on your business needs. Before launching into any big data project, which is really what IoT is all about, you must have a sound understanding of what can and should be collected. Once you understand this, you can seek out optimal analytics software to create valuable intelligence with the data you want to collect.
Much of the concern surrounding the expected explosion of IoT in the enterprise revolves around doubts that many organizations will properly secure data from end-to-end. This is a valid concern.. It's all too easy to start thinking about data security after it after the project is well under way and controls can no longer be properly put in place. Security must be at the forefront of any IoT project. Make sure that your IoT initiative starts and ends with a robust security architecture.
IoT support: skills and logistics
Many IoT projects are going to stretch the geographic boundaries of many enterprise networks. Because of this, you will not only have to plan IoT support from a knowledge and skillset perspective, but also from a logistical point of view. Make sure you can have the right support resources in place before you move forward. And if you can't, an alternative choice is to outsource some of your remote support needs.
Plan a modular IoT architecture
Do your best to design an IoT system that is highly modular in nature. No matter if you choose to be fully-on premise, leverage cloud services, or go with a hybrid approach, modularity is key. Your goal should be to design a platform that can handle all current and near-future IoT needs today while providing a scalable path to reuse and/or easily upgrade the underlying infrastructure when future IoT projects demand it.