IoT Device Management: Challenges and How to Evaluate Platforms

Developing a fleet of connected assets is a game-changer for companies across transportation, logistics, manufacturing, and energy industries.

One of the core challenges to a successful fleet of connected devices is device management.

75% of companies electing to roll out a connected fleet in-house fail to deploy on schedule. Even when they do, device management across large geographic areas is difficult—and even harder to scale from a technical, operational, and logistical standpoint.

Fortunately, companies can accelerate their GTM timeline and manage devices at scale across their IoT deployments by partnering with a fully integrated IoT Platform-as-a-Service.

Let’s take a look at some of the common stumbling blocks for companies developing IoT capabilities in-house, as well as some of the main considerations for choosing an IoT device management platform provider.

IoT devices are Internet-connected, data-transmitting machines, sensors or other physical devices that carry out specific tasks.

IoT device management refers to the broad series of tasks and interconnected functions related to the set up, configuration, and maintenance of these devices. For companies using or planning to develop an IoT solution, these include:

• Device setup. Bringing connected solutions to life requires deep expertise across things like cloud scripting and firmware—not something you’d typically find in a single employee.
• Admin and monitoring of device fleets (and volume management). Once deployed to the field, connected devices require monitoring and maintenance to ensure operation.
• Managing through the device lifecycle. Increasing functionality and features or developing improved uptime are key parts of managing a device through its lifecycle.

For companies with IoT solutions, device failure is a costly proposition.

Consider the idea of remote device monitoring, for example. With IoT, the ability to control devices in the field is crucial. Companies want to be able to verify that devices are connected, that they’re transmitting and receiving data, and that their systems are essentially working—things that, without remote accessibility of devices, require constant physical monitoring to achieve.

But device management in this context also means being able to solve problems from a distance. Sending technicians out to the field to recover devices is expensive, and often impossible. So the ability to monitor devices from afar—and even fix them, in some cases—can save significant amounts of time and money.

For HVAC distributor Watsco, creating connected sensors and monitors to continuously monitor the health of customer assets in the field helps prevent minor issues from going undetected and spiraling out of control. But by selling this remote monitoring capability as a service contract, Watsco was also able to use IoT to generate a new stream of monthly recurring revenue.

Without the ability to guarantee uptime, Watsco would be losing money—and damaging its reputation—every time one of its devices in the field got bricked or went offline.

Configuring remote devices allows companies to iterate solutions at higher volumes and over larger geographic areas.

You can think of effective IoT device management as a company’s ability to build linear processes and solutions that work at scale. In a phrase, it helps alleviate costly repairs.

Companies should think about device management before even building their first devices. Why?

Moving from prototype to production is difficult if you haven’t figured out how things will work when you have a larger fleet.

The two critical factors for device management are volume and proximity. The moment when the volume of connected devices exceeds the number of technicians the company has available to develop and maintain them is the same moment IoT devices become more difficult to manage.

In that scenario, having poor initial design can be difficult to correct at scale. Piecemeal solutions can make fleets difficult to upgrade. Supply shortages such as the global semiconductor shortage demonstrate how difficult it can be to get your hands on the right hardware to build new devices or rework outdated ones.

Another crucial moment is when proximity comes into play. When companies reach scale and begin to ship devices across the country or other large areas, rolling a truck to deploy patches or fixes becomes costlier, impractical, or outright impossible.

When does device design become critical for management at scale? The real answer: it should be from the beginning. But when a company’s production of an IoT product hits a critical mass (e.g. when volume or proximity become an issue) is when the fruits of good device management show.

We’ve touched on the importance of device management and scalability for IoT. But what are some of the common challenges posed by IoT deployments? Some overarching themes include issues related to sourcing, talent development, cost, and access to vendors.

At a more granular level, these include:

• Hardware availability. Sourcing hardware components isn't a major challenge when you have a very limited number of devices in the field. But as your fleet grows, it becomes increasingly difficult to source the tens or even hundreds of thousands of particular hardware components you need. Supply sourcing is one of the major challenges to scaling an IoT product.
• Talent and knowledge development. One of the reasons IoT is so complex is that it spans multiple technical disciplines: hardware design, firmware design, connectivity management, IoT security, cloud infrastructure, and several others. Putting the right people in place to bring together these fundamentally disparate entities—especially on a bootstrapped budget—is a feat in itself. Also, weak design in any of these areas can make your product just that much more difficult to retrofit later. Skimping in these areas is costly for businesses. For example, you need to ensure ironclad cohesiveness between the cloud and the firmware to prevent devices from becoming remotely accessible.
• Cost. For many companies, the total cost of ownership for IoT projects poses a significant financial barrier to entry. But creating truly resilient devices also requires trial and error—something that pushes out the total timeline for IoT projects. Building internal capabilities also requires spending for technical expertise in specific areas.
• Access to vendors. For companies with small deployments, it can be difficult to negotiate favorable terms and meet order quantity requirements with carriers and other vendors. Even at scale, companies can struggle to get the support team to resolve their issues efficiently.

The above stumbling blocks show just how difficult it can be for companies to build out their own fleets of connected devices. It’s why many look to speed up their time-to-market and make their products more easily reprogrammable and retrofittable with the help of the right solution.

But choosing the right platform for your specific use cases and needs can be difficult.

Here are some of the most important features to keep in mind when evaluating device management (and, more broadly, IoT), solutions:

Remotely Controlling IoT Devices

Remote access means you should be able to monitor and control your devices and assets from anywhere in the world. When you control devices remotely, you can send data and requests to communicate and check functionality.

Being plugged into a real-time data interface lets you adjust settings and detect bugs before they get the chance to damage your clients’ systems or devices.

Sending Over-the-Air (OTA) Updates

The ability to send OTA updates to your devices in the field means you can totally reprogram your devices at any time. A good device management solution must include a robust OTA solution as a core component.

Code isn’t static—you’re going to deal with bugs over the lifecycle of your product fleet. OTA updates make it easier to deal with them efficiently. For IoT-enabled hardware products, that means you can get them to market and then catch and manage bugs post-production.

It’s also likely that you or your client will want to add new features and capabilities to a currently deployed fleet. With OTA updates, you can remotely upgrade capabilities from anywhere.

(That’s why Particle excludes them as part of your company’s data operations limit. That means you can update your fleet without incurring extra cost.)

Billing on Data Operations Versus Data Consumption

Speaking of…

Particle opts for a transparent and simplified model based on the number of data operations you process as opposed to data consumption. But why is that good for you?

Data-consumption based models charge based on the amount of data you actually use. This sounds great in theory—but the problem is devices send different-sized data payloads all the time. (And traditional data models charge for data used, even if it’s dropped or doesn’t make it to the device or to the cloud.) This means that costs are harder to predict and control.

On the other hand, our data operations model charges you one flat rate for calling your device. This means you can call your devices and get data no matter what. Particle doesn’t bill on bytes—just by actions.

Monitoring Device Activity and Health

There’s nothing worse than not being able to contact your cellular provider when your fleet goes down unexpectedly. When you partner with an integrated IoT Platform-as-a-Service like Particle, you get all the benefits of real-time status reporting combined with the level of access to the cellular companies that only large aggregate vendors enjoy.

Particle can analyze large patterns of data across all of our customers to inform you of an outage before anyone else, or combine your company’s anonymized usage data with others to tell you the size of the outage area.

Our network plugs us into data insights that the cellular companies don’t provide—like when we tracked the sunsetting of 3G services and found it didn’t correlate with the dates the telecom providers themselves gave.

Connecting with Webhooks and Integrations

Device data often needs to be routed to different systems and third-party platforms to be useful. Your device management solution should allow for easy integrations with common platforms.

Particle offers first-class web hook integration that allows your devices to seamlessly interact with outside web services from cloud platforms like Microsoft Azure and Amazon Web Services to BI tools like Salesforce or NetSuite.

Why makes Particle’s offering unique? It starts with our approach to IoT protocols. Particle employs a CoAP instead of an HTTP protocol, which allows us to efficiently transmit data down to devices in even incredibly constrained or low-connectivity environments.

But because HTTP is the language of the web, Particle also handles all protocol translation and lets you talk to your devices over standard HTTP web interfaces from your own services.

With Particle, you get the best of both worlds—where optimal connectivity meets frictionless web server integrations.

Making Cellular Connectivity Easier

Reliable connectivity management is a must-have in any good IoT device management solution. Cellular IoT connectivity is one of the hardest things to get right when building a connected product.

Particle pre-integrates SIM cards directly into your hardware devices. The upshot here? You get fast, reliable and global cellular coverage—all without ever having to deal with SIMs or MVNOs yourself.

Facilitating Scalability and Organizability

Hardware devices are tricky to deploy, difficult to source, and a beast to retrofit later. Particle’s integrated and unified PaaS means you get the hardware you need to grow your fleet without the growing pains of bugs, shortages, and the problems other platforms create.

Integrating the Operating System into the Cloud

Your devices should have a functional and integrated IoT operating system that’s directly connected to the cloud. If your vendor doesn’t have that, you have to hire costly cloud and firmware engineers—and then rely on them to bring everything together.

But unlike other platforms, Particle is truly integrated. We help you design and build the cloud infrastructure and hardware devices you need from the jump—meaning things just work.

You don’t need to build your IoT product with a hodge-podge assortment of vendors and homegrown solutions. Partner with Particle once and enjoy a single solution that combines the entire stack of IoT technologies.

Building Deep Relationships with Cellular Carriers

Your platform should take the work of dealing with cellular carriers off your plate.

Unlike other IoT PaaS solutions, Particle negotiates with cellular vendors on your behalf. You get immediate access to tier 1 carriers without having to juggle multiple contracts.

Particle’s relationships with major cellular carriers also means that we can work with them more efficiently to resolve issues around device connectivity and ensure the highest level of uptime.

One of the benefits that this level of access provides is an extended NAT timeout.

It works like this—your device will need to constantly send data to stay connected to most cellular networks. That means you’re paying just to stay connected, even if your devices aren’t sending useful data.

Our NAT timeout is 23 minutes—meaning your device only has to send data for that interval to maintain bi-directional communication. Particle’s bidirectional connectivity is low-power and less resource-intensive. And for comparison’s sake, the average AT&T NAT timeout is 60 seconds—requiring much more power just to stay connected.

If three-quarters of companies that attempt IoT device management on their own fall behind schedule, it’s a clear sign that getting it right is hard.

IoT device management comes down proactively planning for the challenges we’ve discussed here by identifying what you need, when you need it.

And one of those needs is working with the right IoT partner.