5 Hurdles to IoT Device Development and How to Overcome Them
If you’ve attended a technical tradeshow in the past five years, odds are you’ve stumbled across a slew of new devices being marketed as a “Smart This” or “Connected That”. NAIAS, Neocon, RVX, NBAA, and certainly CES have all featured innovative products centered around the new industry we’ve all come to know and love: the Internet of Things (IoT). If you haven’t attended a tradeshow (or aren’t technically inclined), odds are you’ve still experienced this cultural shift. From refrigerators that talk to your phone to Amazon Echo (Alexa) and Google Home, the ethos of today’s products is largely centered around connected convenience. We want all information to be accessible with the touch of a button or through a short voice command.
This drive towards connectivity has forced major changes within product development for organizations. Consider this: a company that has designed and built solid-oak desks for years and years suddenly needs to include a USB port or wireless charging into the design to facilitate connectivity. A home door lock now needs to connect to the internet to provide remote locking/unlocking and connect to a phone over Bluetooth, so a door automatically unlocks when an authorized person approaches. Suddenly, companies that have been one-trick ponies for hundreds of years need to learn how to design PCBs, write firmware, get their devices connected to the internet, and maintain security from hackers. This can be quite a minefield to navigate for resource-constrained companies that suddenly need an experienced electronics group overnight.
As a member of DISHER’s electronics team, I have witnessed this challenge firsthand with multiple customers. If you are an organization heading towards IoT product development in order to stay relevant, there are a few common misconceptions to overcome with your first smart-device project.
5 Hurdles to Overcome
1. Cost Expectations
The most common misunderstanding is a misalignment between what a company new to electronics development thinks a project should cost, versus what it really takes to develop a new product. Many companies come in with the expectation that the electronics will cost approximately the same amount as another piece of a typical mechanical project pie. Speaking in broad generalities, the electronics development on its own is every bit as large as a standard mechanical project. There are pieces to an electronics project pie as well, all of which add up to a full additional project on their own. (See Figure 1 and 2).
2. “It’s just software”
If DISHER charged a nickel for every time a customer uttered the phrase, “it’s just software”, we could give the Fortune 500 a run for their money. This ties closely to misaligned cost expectations. There’s a fine line between a task that is very simple and a task that is nearly impossible. The common sentiment is that all software is “just software.” Since there’s no physical deliverable, the assumption is that the software comes preloaded on the PCB or that creating the software is as simple as clicking a few puzzle pieces together. In many ways, it is like having the expectation that when you purchase a house—you should get all of the furnishings included for free. The reality is that filling your new house with the furnishings you want can be expensive. In some cases, the appliances, furniture, lawn equipment, systems, and decorations can rival the price of the house depending on the quality and complexity. My experience with this thinking is summarized quite well by Randall Monroe, author and webcomic of XKCD.
3. V1.0 Release
There’s a phenomenon that’s all too common in new products (not exclusive to electronics). The team wants the initial production release to be absolutely perfect. In the software world, this means completely bug-free and with a flawless user experience. While this desire is understandable, nothing ever launches as a perfect product. Sure, there is a certain level of polish that’s to be expected. And many bugs should be caught and rectified in the test phase of a project. But even the iPhone, the pinnacle of engineering for many consumers, occasionally has glitches. So, how do you decide when a project is ready to release? This depends on the project itself. Tight agreement to a set of measurable standards and testing to these standards is key. These standards should acknowledge the presence of bugs upfront and set an expectation that the device should fail predictably and gracefully when failures do occur. (Remember, “predictably” and “gracefully” need to be measurable too.)
Some companies have made one foray into electronics and now want an update to add a new feature. Updates can quickly fall into the same “it’s just software” trap, depending on the new feature requested. However, projects in this category have other pitfalls as well. Sometimes the source files have been lost, but Gerber files, pdf schematics, and/or the BOM are still available. To update these, the design effort will need to be essentially started from scratch. While we can leverage the parts and thought that went into the original design, the source files need to be recreated before any new features can be added. Let’s go back to our home analogy. This is like going to a contractor with blueprints of your house and asking them to build an addition. However, the house burned down last year. In order to get your addition, you’ll first need to rebuild the house. The thought (the blueprints) can be referenced, but it still takes time and materials to build your house back up.
5. Maintenance & Security
Another hidden cost to an electronics IoT project is maintenance. Once a project is complete, how do you make sure that the bugs found in the field are addressed? If you build a device that relies on Amazon Alexa as a voice assistant, what’s your plan if Amazon decides to eliminate the whole project, taking the backbone of your device and rendering it useless? Designing your product is just the first step. Once it’s out in the wild, there’s a whole new area of your business that needs to be dedicated to keeping the product current and safe. Cybersecurity is a real risk to all connected products and requires a team dedicated to supporting your new product.
In summary, there’s clearly a lot to consider when determining whether or not to include a PCB inside your latest product. There’s even more to think about if you want it to connect to the internet. With nearly a decade of electronics projects under our belt, DISHER has the experience to help overcome these and other obstacles to successfully bring your next IoT device to market. DISHER would be excited to partner with you on your first (or hundredth) electronic device.