IoT Blog
IoT Blog

Industrial IoT is Freaking Hard

by Olivier Pauzet, Vice President & General Manager, IoT Solutions
When people ask me why Sierra Wireless introduced Octave, the new all-in-one edge-to-cloud solution for the Industrial IoT (IIoT), I tell them a story. It is a story about how the IIoT might seem easy, but, when it comes to actually scaling an IIoT application proof of concept (POC) to commercialization, it can be hard. Really freaking hard.

The story comes to me naturally since I have heard dozens of versions of it from Original Equipment Manufacturers (OEMs) and other customers over the years. 

The story starts with you (or someone else at your company) thinking up a great idea for a new IIoT application. The new application will connect your company’s commercial washing machines, air compressors, refuse bins, or other industrial equipment to the cloud. By collecting this equipment data, your company will be able to create a new equipment monitoring, predictive maintenance, equipment-as-a-service, or other type of IIoT application. With this application, your company will become part of the movement towards Industry 4.0, using the IIoT to transform its business by improving equipment performance and uptime, reducing equipment maintenance costs, or offering customers new consumption-based services. 

A Promising Industrial IoT Proof of Concept

You tell your boss about the idea, and she is intrigued. She signs off on you building a POC of the IIoT application. Your task now is to build a POC that will show your boss and your company’s board that this idea is not pie in the sky, but something you can really do, today.

To connect to your equipment, you will probably use an open-source IoT development platform. If you use something like Raspberry Pi, building this connection will probably take longer than you expected and not easily provide you with all the functionality you want, but you will eventually get it done. Alternatively, if you use Sierra Wireless’s mangOH open source hardware platform, connecting to your equipment will be easy and quick, and allow you to immediately start collecting all the data you need for your POC. 

You also set up an account with Microsoft Azure, AWS, or another cloud provider, where you can build the application that will collect, store, and use the data collected from your equipment. After a few fits and starts, you have data from your equipment being transmitted to the cloud, and the POC up and working.

It is looking great! Your cloud developer team has built an application that can extract the data you wanted to collect from the equipment. The application itself, while still rough around the edges and not ready for “prime-time”, demonstrates that it is possible for you to monitor your equipment’s performance, predict when it might need maintenance, track its usage – whatever you need for the IIoT application’s use case.

You present your IIoT POC, along with a preliminary financial analysis of the costs and ROI of the IIoT application, to your boss and your company’s board. They are impressed. Finally, the company will be able to use the IIoT to make its customers happier, to reduce maintenance costs, or even to create new revenue streams. You are given the go-ahead (and budget) to begin commercializing the POC, with the caveat that they want an update on progress in six months. After a few congratulatory pats on the back, you leave the meeting, excited that your company will be part of the new industrial revolution with an IIoT application that will transform the way it does business.

Industrial IoT Application Commercialization - Slow, Complex, and Risky

So, now that you need to make this POC real, what first? Well, security of course. When you were just building the proof of concept, basic security was fine – your application wasn’t a visible target, and the worst that could happen was someone accessing your test data. But now you will be collecting customer data from devices located around the world and transmitting that data into your or your customers’ Enterprise Resource Planning (ERP), Customer Relationship Management (CRM), or other systems. You need to make sure your IIoT’s application’s security is rock solid. But this shouldn’t be too much harder than making a cloud application secure – right?

Wrong. The IIoT brings with it a whole other set of security complexities, above and beyond those of cloud applications. Your industrial equipment is often used in accessible, unsupervised locations, which makes it easy for someone to tamper with the edge devices connected to this equipment without being detected. Also, your edge devices have limited bandwidth and battery power, making it more difficult to deploy security updates in a timely fashion. Finally, finding someone with the deep levels of IIoT security expertise needed to do all this – as well as implement IIoT security best practices that you are not even aware of – can be a nightmare. 

Well, at least connecting your edge devices to the cloud will be easy and straightforward. Or, at least you think that until you try to work with the Mobile Network Operators (MNOs) in the countries your equipment is used. They all have different, complicated data pricing plans, and for every network provider, you need a different management platform and SIM to put in your edge device. The MNOs’ complex pricing and the additional work required to install and manage all these different SIMs has made you a lot less confident in your ROI forecast for your IIoT application.

In addition to dealing with the complexities of managing connectivity in different geographies, you have also hit a roadblock related to the different needs of your customers. One of your key customers likes your IIoT application POC, but will only deploy it if you add some other capabilities to it. When you start to work on updating the application to provide these customer-specific capabilities however, you learn it will require you to rewrite the edge device’s code from scratch. In addition, you need someone with deep levels of embedded device expertise to make this change – and finding that person is just as hard as finding the IIoT security expert. 

These are just some of the roadblocks you encounter. On top of this, add the challenge of trying to find a way to filter and process data at the edge device, to ensure your data transmission charges stay low, and your edge devices’ batteries don’t quickly exhaust themselves. You also need to rewrite your cloud APIs when you determine you need to switch to another cloud provider for the application, and write new APIs from scratch so you can connect to your own or your customers’ ERP or CRM cloud services – and then get the right IIoT data to these services at the right time. Plus needing to scale the application from one edge device to thousands of devices located around the world, and find a way to monitor, control, update, and maintain all these edge devices and their SIM connectivity. You get the picture.

What you have discovered is that to actually commercialize your IIoT application, you need to build your own IIoT infrastructure from the ground-up. Not only is this slow, complex, and risky, but it is distracting you from the whole point of the project – to build an innovative IIoT application that delivers your company the data it needs to transform its business. 

You meet with your boss and the board for your six-month check-in on the IIoT application. You inform them that you have missed most of the deadlines you set for commercializing the application. That you think you need to hire specialists in IIoT security and embedded device software – assuming you can find them. That your edge device connectivity costs are both higher and less certain than you expected. That you have had to significantly lower your ROI estimate for the IIoT application – and you lack confidence even in this revised estimate. You leave the meeting to wan smiles, but you know what you will hear from your boss tomorrow. The board has pulled the plug on your IIoT project, and like approximately 75 percent of other IoT projects, it has failed. 

Octave Makes the Industrial IoT Fast and Easy

This story is the reason we developed Octave. To provide you with a solution that has end-to-end IIoT application security built-in. To offer you transparent, predictable connectivity pricing, along with the ability to work with one connectivity provider, with one SIM, for global IIoT deployments. To make it easy for you to create different versions of the IIoT application for different customers and update your IIoT application as your business needs change. To make it efficient and easy to deal with constraint devices with low power requirement. To make it simple for you to process and filter edge device data, integrate this data into different cloud services, and monitor, manage, and control all your edge devices. To deliver you IIoT infrastructure and software APIs that – like Google Maps, Tableau, Stripe and other popular API-based services – plugs right into your application and just works. 

The result? You can simplify and speed IIoT application development and commercialization. You can de-risk your IIoT application deployment, and you can actually focus on your data, not your IIoT infrastructure. The IIoT does not have to be freaking hard. In fact, with Octave it can be fast and easy. 

Start with Sierra to learn more on how we can help you simplify the Industrial IoT, with everything you need to bring your Industry 4.0 ideas from pilot to deployment.