If you’re a business preparing to implement the Internet of Things (IoT) into your operations, it’s likely that you’ll have already considered a number of options and opportunities to prepare for a connected device deployment. After all, Machine to Machine (M2M) IoT is a complex area and each option can easily generate further questions, resulting in an overwhelming number of possibilities. However, the most critical is arguably: how do you ensure your project is a business success throughout its lifecycle?
Amidst so many new supporting services and developing projects, rapid global adoption has ensured that IoT continues to grow very quickly, To ensure that your business gets its deployment right in the long-term, it is essential to evaluate both your IoT strategy, and your capability to execute it. Unfortunately, for those businesses that progress without conducting a thorough analysis, IoT deployments can produce a number of issues and complications.
Here is a small selection of the critical questions you should ask before implementing your IoT plan however, it is more than likely that more questions will arise as your project develops:
1. What are the most effective options for device connectivity: cellular, wired, Wi-Fi or radio?
If a wire or WiFi, do you have control over the setting in which the device will exist? Do you know how much the installation of network cabling or infrastructure at each location will add to the cost of your project? Do you control the WiFi? If you don’t, but link to it, how will you ensure you stay connected at each location when something happens on-site? If Radio and you’re deploying worldwide, how will you ensure you adhere to local legislation regarding frequencies, transmitter powers and the approvals you’ll need for each country? If Cellular, to what extent can you be clear about where devices will be located and whether they are fixed or moving? A general rule is: the more global and the more areas your devices are entering, the more you’ll need to invest in the planning and optimisation of the device.
2. What is the financial cost of down-time and lost connectivity?
We advise you to calculate this carefully, and we can help. You might be surprised at the speed in which the numbers escalate. For example, single network operators average around 80 per cent coverage at the locations in which IoT devices are installed. The more downtime costs you, the more you should invest in the build and in connectivity options. You should also estimate the cost of onsite maintenance visits and do everything you can to minimise these. Early procurement savings may have long-term negative results on the project if ongoing costs escalate.
3. How can I build devices that will stay connected, wherever they are deployed?
Do you fully understand the range of environments the device is expected to work in? This knowledge is crucial and has significant impact on the hardware and software choices you will make. Will you go inhouse or use a third-party contractor for design, and how much are you prepared to spend on device hardware? Eseye has over ten years of device optimisation experience on more than 200 projects, and also manufactures a range of high-end Hera routers, so knows well that there are pros and cons to both approaches. This is another area where early planning and investment will reap rewards further down the line.
4. For cellular connectivity devices, how should I contract and connect them across multiple locations, regions and countries?
If you plan to manage this inhouse, you’ll need to build expertise and capacity. For global deployments you’ll need to understand local regulations, which for example, can stop a standard roaming SIM card working across a range of countries. Nor is the legislation as light or as simple as you might wish: different countries require different certification and are subject to local laws. In terms of contracting with Mobile Network Operators (MNOs), if you engage in a multi-country deployment, be prepared to find their world fragmented. There are around 800 MNOs worldwide and individual contracts will result in you having to manage across multiple SLAs, languages, customer support centres, data price points and billing formats. Plus, if your business case requires over 80 per cent network coverage, you’ll need to contract with at least two different network operators in each country. There are also in-country roaming rules you need to keep on top of. An alternative is to engage with companies like Eseye, who act as a single point of contact and contract and manage all of the above.
5. With the assistance of customer support services, how can I make sure I get problems fixed quickly?
When you have issues, and there will be some, resolutions will be quicker and more successful if you can talk direct to IoT experts who understand the device and the way it might operate on the network. Not all MNOs deliver this, so make sure you check and get references on customer support from existing clients with similar scale deployments to your own.
6. What type of cloud storage is the best fit and how will I make sure valuable data can be reported and analysed effectively?
This is the essence and value of IoT: to be able to gather information from all devices and centrally view and manage them, to save or make money, or to deliver some other kind of impact, such as social or environmental. There will be a budget and benefit analysis to undertake, but you have two main routes to choose from: either build inhouse server and developer capacity or buy-in Cloud services and configure your project around their rules. Your requirement may be so unique you have no choice but to retain a developer team and to build and manage inhouse hardware and analytics software. However, for most, customisable off-the-shelf storage and analytics solutions will be fit for purpose and give better value and scope to scale. Bear in mind that the bolt-on data analysis and security device management software that companies like AWS are developing will continue to add services; just as long as you are willing to configure by their rules, e.g. MQTT (Message Queuing Telemetry Transport). On this note, in establishing their rules, cloud providers have invested thousands of hours designing, building and testing for reliability, in the wide variety of real work scenarios. Whilst you may initially find adhering to some of the design rules and standards difficult and will be tempted to design a simpler custom solution, think hard about whether this is a sensible and scalable, long-term strategy.
7. What steps can I take to ensure my deployment is as robust as possible throughout its lifespan?
It is essential to consider the implications of someone getting into your system before it happens, and to invest at the right level. We advocate and provide a range of advanced security features, which bring new, more affordable protection, and at the same time adhere to the security requirements of AWS Cloud. As you prepare your devices and connectivity you should decide where your security boundary is and plan accordingly. Give serious consideration to contracting in third-party penetration testing experts and ensure your team is keen to encourage this.
The deployment of an M2M IoT solution is often highly complex, particularly for global deployments. In the process of doing so, your business must therefore consider many components to ensure its needs are matched at an affordable cost. However, by taking the time to ask the right questions up front, you’ll be far better placed to find the right provider to serve as a single resource for your M2M needs.
Ian Marsden, Founder and CTO, Eseye
Image source: Shutterstock/a-image