Commercial IoT is on the rise. Designing an IoT solution for a Smart Office Building, Smart Retail, Connected Guest Experience, Location and Asset Tracking or similar application is a challenge being put in front of more project teams every day. These requests are growing at a feverish rate based on the proliferation of low cost sensors and beacons, the affordability and elasticity of the established cloud computing market, and a growing big data trend. This combination of technology, market trends, and affordability has paved the way for a new class of connected solutions, Commercial IoT, which differs significantly from Industrial IoT and Consumer IoT in several critical ways.
Naturally, project teams being put on Commercial IoT initiatives are pulling design ideas and technologies from their more established market neighbors, by borrowing technologies, solutions, business models, and design ideas from either Consumer IoT, Industrial IoT, or both. This borrowing of ideas is a chief cause of the high IoT initiative failure rate, as solutions for those markets are not specifically designed for and tailored to the unique challenges of Commercial IoT. According to a Cisco study, 60% of all IoT initiatives stall at the Proof-of-Concept stage.
In this white paper we will discuss the factors surrounding the creation of IoT solutions, and how those factors differ between Industrial, Consumer and Commercial IoT. We’ll conclude with a discussion of the features that an ideal solution might have in order to successfully overcome the unique design challenges of building a Commercial IoT solution.
Rigado has been fortunate to work with hundreds of companies building and deploying large-scale Commercial IoT applications. We began helping companies design and deploy Bluetooth Low Energy (BLE) solutions in 2010 and have grown with BLE as it expanded from early Consumer devices to large-scale Commercial IoT applications. Our goal is to help teams avoid common design pitfalls through an analysis of the unique design challenges presented by different types of IoT solutions.
Consumer IoT, which includes things like wearables, monitors, and smart home applications, typically relies heavily on a central mobile device. The smart phone, tablet, or media hub at the center of consumer solutions provides a rich environment to simultaneously connect to devices and run multiple applications. The primary device connectivity choice is two-way Bluetooth, with just one or a small number of simultaneous end-device wireless connections and the mobile device acting as the gateway to the cloud.
These mobile devices have significant processing power, comparable with PCs of just a few years ago but in a significantly smaller hardware footprint. Thanks to their ubiquity, applications are relatively easy to build, storage space is rarely an issue, and there are a plethora of engineers with an understanding of the environment. Security is largely controlled by the operating system and, in a significant way, by the user’s ability to keep track of and password-protect access to the device. Software can also be employed to ward off unwanted viruses, data theft, and unwanted, external network access. While the initial cost of a consumer solution is taken into consideration, it is not ultimately the determining factor for purchase – if the consumer really wants it, they will find a way to get it.
It’s easy enough to imagine building out a mobile-device- centric solution for Commercial IoT (i.e. mounted in the ceilings of a Smart Office Building). One might even think of a Smart Building as being a group of Smart Homes sitting very close to, and on top of each other. But this mobile- centric Consumer IoT design approach is not one that scales well to large-scale Commercial IoT deployments. Scaling such a solution would be too complicated, would create unnecessary system vulnerabilities, and would be too expensive to create the necessary ROI.
Industrial IoT (IIoT) is the oldest and most mature of the IoT markets, as it evolved out of production monitoring and maintenance procedures that have been industry standard for many years. Recent advances in the IIoT market have come by way of big data analysis or artificial intelligence, which enable automated and highly-informed decisions at the edge. IIoT use cases cover mostly machine-to-machine communications that monitor, and control large, complicated, expensive equipment used in industrial processes such as energy creation/extraction, factory floor manufacturing, or any other large-scale processing operation. IIoT installations often use a mix of hardwired (e.g. RS-232) and wireless connectivity. Connections are usually one-to-many between sensors and a gateway device (to provide sensor connectivity) and include some level of local command and control. The data being transferred is very different from that of Consumer IoT, and is often categorized by type, such as temperature, location, vibration, etc. Many of these solutions are AC-powered, thanks to easy access to power from the factory floor or on a piece of heavy machinery and where expensive, server-class gateways are the norm. Budgets for IIoT solutions are often very large due to of the value of the machinery (or uptime value of the machinery) that is being monitored.
These IIoT considerations often lead to solution design patterns in which functionality, resilience, and reliability can be maximized, and where there may not be significant constraints when it comes to battery life, size or support costs.
As alluded to above, the growth of Commercial IoT has been fueled by two accelerating trends. First, there has been a proliferation of affordable short-range wireless devices – using protocols such as Bluetooth Low Energy and 802.15.4 – with projected growth to 16 billion devices by 2022. The second trend is the rise of edge computing, in which gateways connect these wireless devices to the cloud and run applications to collect data, analyze data, control devices, reduce data, and deliver new user experiences at the edge.
Commercial IoT, as the name suggests, is often used to categorize an IoT product or solution that is about creating commerce, that is about augmenting business to improve profits either by adding a new revenue stream or by reducing costs in current operations. An example is an IoT occupancy sensing solution for conference rooms in commercial office space. This solution is sold as an added value service by the building owner (additional revenue stream) and is used by the tenant to improve employee efficiency and workplace happiness.
By this nature, Commercial IoT solutions have a strict requirement to show ROI. As such, the solution designs are typically very cost sensitive.
Commercial IoT solutions are additionally characterized by a need for wireless sensors/end devices that can be installed relatively easily, moved if necessary, and can run effectively on battery power for long periods of time. Enough end devices are required for data collection to necessitate an edge router or gateway device that not only sends the aggregated data to the cloud, but that can also run edge applications for local command and control. Since the sensors are wireless they are often used in an intermittent connectivity mode – this both reduces the power consumption of the sensor or beacon (which is critical in Commercial IoT) and simultaneously also allows a single gateway to monitor more of them, because the connections are one-way and not continuous. Standard Bluetooth, therefore, is not ideal due to its high-power consumption and the fact that it does not scale to enough simultaneous connections to be viable. Bluetooth has a low energy version (BLE) that solves this issue. BLE is increasingly common in Commercial IoT solutions, but there are also several other popular options like Thread, Zigbee, and Z-Wave.
Also unique to Commercial IoT are several environmental challenges that you do not usually encounter in the other two markets. In many cases, such as an office building or warehouse, there are great distances between the devices and gateways. Further, there are often walls or other barriers that can impede wireless connectivity over these distances. And finally, there is a possibility of being outdoors, requiring that the gateways and devices be able to withstand harsh weather (moisture, extreme temperatures, etc.).
High sensor/device quantities requirements, internal and external environments, and a need for cost effectiveness are making mesh network connections using Bluetooth Mesh, Wirepas, etc., an increasingly popular choice. Mesh networking allows for sensors and beacons to cover larger areas with fewer gateways, because those end devices are able to network with each other to extend their range beyond a standard gateway-to-end-device connection. Because Commercial IoT solutions are used in private commercial or enterprise-like solutions, even less common or proprietary protocols are more likely to find a place. Unlike Consumer IoT, interoperability and consistency among wireless solutions is not as big of a hurdle, and each protocol offers features and efficiencies that gives it a fighting chance in solving a challenge in a more effective way.
As you can see, there are quite a few differences between the requirements for a Commercial IoT solution and those for Industrial or Consumer. The next step is to figure out, given these differences, what the design of the ideal solution would look like. What characteristics would it have to overcome the challenges associated with these differences and create an optimized solution? What we can determine is that the typical effective Commercial IoT solution will require, as its core, an edge computing capable, flexible, affordable, secure, hardware-optimized gateway. The gateway must allow for a variety of device connection protocols to work simultaneously and at scale. It should also easily connect to public or private clouds and be able to run multiple applications that have been optimized for edge computing. The gateway plays a critical role in the Commercial IoT solution design, and significant evaluation time should be spent on finding the “right one”.
So, you’re going to need a gateway that does these things, but how do you decide on one? As we have discussed, borrowing from Industrial IoT solutions design is a bad idea, as those gateways are unnecessarily complicated and expensive. Using a cellphone or tablet (borrowing from Consumer IoT) is also a bad idea, as it cannot scale, is vulnerable, and is also too expensive.
The next logical choice might be to build your own gateway using something like a Raspberry Pi modified with some extra connectivity modules and antennas. This choice is appealing because you can get started very affordably and you can customize your setup to include that right mix of power, storage, and RAM. The downside to this idea is that a Raspberry Pi solution cannot scale. Can you imagine trying to manage 100, 1,000, or even more of these as gateways in a Commercial IoT solution? The simplicity of the POC becomes the breaking point at scale. You would need to deploy, manage, update, and secure in mass and frequently. Further, you would need a full development team with an expertise in each of these challenge areas. For a commercial organization whose main business is not technology, but rather commercial real estate, retail, hospitality, shipping, or similar, that is a large and costly undertaking that is unlikely to be ultimately successful.
Taking everything that we have discussed into consideration, the ideal solution for a Commercial IoT project would be a device-to-cloud connectivity platform, including gateway hardware, that makes it easy to deploy, manage, secure, and scale. The gateways would connect to many end-devices via a variety of different protocols, but remain simple to setup, develop for, and use. The whole system should be secure from end-to-end (device-to-cloud) and be automatically kept up-to-date and protected as new vulnerabilities or exploits are found. To be a great solution it should be flexible enough to allow you to develop and seamlessly update your own uniquely differentiating applications (for Smart Buildings, Smart Retail, etc.) while doing all of the rudimentary, technology, and security “heavy lifting” for you. The ideal solution would allow you to start with little or no investment during the POC phase, and then scale the cost as your solution scales in use and features, even to the highest levels of scale found in Commercial IoT. This specifically designed edge computing and connectivity platform would be the optimal solution for Commercial IoT and would offer the highest possibility of initiative success.