Four IoT Lessons from the Sierra Wireless Innovation Summit

At the Innovation Summit 2018, important practical realities about IoT were brought into the open. The first is that cost concerns about Low-Power Wide-Area (LPWA) technology are overplayed and that the race to the bottom in module and connectivity pricing can only go so far toward enabling a market. Second, the heterogeneous nature of the “edge” prevents IoT device data extraction from being as simple a process as might be assumed. Third, scaling an IoT project beyond its proof-of-concept (POC) stage can unexpectedly reset the entire operation, resulting in project failure. And last, despite the software- and developer-centric world that we now live in, there are purportedly less than 1 million embedded developers worldwide—a critical enabler and, in the light of this, a critical obstacle to the growth of the IoT. Each one of these issues is worth exploring in more detail.

Like Sierra Wireless, most cellular module manufacturers are pinning their mid- to long-term revenue expectations on Cat-M and NB-IoT; specifically, on the shipment of a large quantity of Cat-M and NB-IoT modules and, more importantly, on the leasing of managed services to maintain and control an enterprise’s network of connected devices. Cat-M and (especially) Narrowband Internet of Things (NB-IoT) are technologies designed to network endpoints that are too numerous, remote, or individually low in value to have been economically connected before. The modules were always intended to be highly affordable, but the issue of price has become overemphasized. There comes a point where, if a viable IoT business model cannot be built out of a material cost of a few dollars to connect each endpoint, then fifty cents, or a dollar, or two dollars less will not make any further difference. Even if modules were free and could be put into everything manufactured, an IoT deployment would still fail without a valid business model behind it.

IoT endpoints are all the same in that they exist to generate events that can be logged as data. But they are also infinitely varied—in form factor, physical situation, power consumption, communication protocol, processing power, mechanical operation, value, purpose, mission-criticality, and industry. Some IoT devices are purpose-built as connected and some may be retrofitted for connectivity, while others may be assets inherited through acquisition. While the process of networking assets is common, individual instances of asset command and control may be completely custom. This needs to be considered up front as a vital part of the business model before connectivity is deployed to avoid ending up with a functioning but useless store of information. Consideration for the technological operation of the IoT system must come later and be subservient to the logical operation of the use case.

Enterprise survey results indicate that that three quarters of IoT projects fail from a technical standpoint and that only half of those that remain succeed in a business sense. It is not bad per se if POCs fail, as their purpose is to determine whether a full-blown project is viable or not. What matters most is whether a proposed project fails for the right reason—i.e., because there is no meaningful Return on Investment (ROI), not because it was incorrectly planned. A key point of failure is the ability to scale. Solutions to a problem can be successfully built, but if those IoT systems need to be redesigned in between graduating from the test bed and entering commercial operation, then the time and money already spent is wasted. This increases the necessary ROI (or extends the time to ROI) and undermines management confidence and therefore budget allocation within an enterprise. As a result, IoT projects have been put on hold or completely abandoned due to a lack of thinking big and focusing on the long-term requirements from the start.

Since ex-Microsoft CEO Steve Balmer’s “developers” chant in 2000, the importance of software developers has been publicly cemented. Hardware and operating systems live or die by their software support, and software platforms live or die by their developer engagement. Yet it is thought that there are less than 1 million developers worldwide specializing in embedded applications. As a point of comparison, there are believed to be more than 9 million developers dedicated to Android and iOS, which is almost half of the global total. Developers will gravitate toward platforms where there is the most perceived return. In turn, the platforms that gather the most developer support become de facto industry standards and, therefore, are the ones most likely to be chosen by enterprises for their projects. The strength of an ecosystem engenders confidence in the parties looking to employ a system, and that brings us back to the common topic of the need for more transparent proof of value in IoT in order to grow the developer community.

Sierra Wireless’s annual Innovation Summit is a gathering of this long-standing cellular module manufacturer’s partners and customers. To call Sierra Wireless a module manufacturer is to underplay the company’s strategic acquisitions and the research and development the company has undertaken in the last five years to move to a service-based business (as it is now helping its enterprise customers to do) as well. At present, the bulk of Sierra Wireless’s revenues still come from the shipment of terminal hardware. The Innovation Summit is a forum for the members of Sierra Wireless’s IoT ecosystem to share and discuss their needs, experiences, successes, and challenges in the implementation of IoT projects and to seek out new business opportunities for their products and services.

It might seem obvious, but a clear vision with proven value needs to come first in the IoT. Companies have spent large sums of money getting caught up in the excitement of the technical exercise of implementing an IoT project. Testing and selecting the right technologies and working out how to implement them, without considering first and foremost what it is that they are trying to accomplish, and what it is that they want to do with the data they have collected is foolhardy. There are companies that have built technically successful IoT implementations that do not contribute toward their bottom line because they do not use the gathered data. These companies do not know what to do with it and are only collecting it because they can. They have no idea how to use the data to help in the running of their business. They are consequently abandoning projects and writing off all that expense, effort, and learning.

Having a specific pain point to address and knowing conceptually what it is that one is striving to change and accomplish is the most critical factor to any project, and IoT is no exception. A valid IoT business model is one where data is at the heart of the value proposition—not one where connectivity is simply built in as a feature or differentiator. Planning for scale and thinking big from the start is required to create a system that is predictable and repeatable and that has adequate capacity. It is not enough for a system to work technically at the onset—it must be architected and guaranteed to work at n times the demand, as this is what constitutes the difference between a science project and business transformation. Technology is an enabler. It serves the business model. Technology should not be rolled out just for its own sake, nor should vendors and service providers push it because it is the latest thing. But when the proof of value is confirmed, the use and purchase of the technology will follow as a natural no-brainer.