4 Essential Steps to IoT Device Development


There was a time when you could just build a smart device, then sell it and forget about it. Connectivity has changed everything. Now, there are more options for developing, managing, and selling an IoT device. Device developers drive innovation, accelerate IoT evolution, deliver new use cases, and continually improve product quality. All of this taps into a major opportunity as the IoT market continues to grow rapidly.

The developer journey looks familiar, whether for sensors, industrial devices or edge nodes, made possible by advances in mobile connectivity and improved components.

For each case, IoT device manufacturers need to come up with a business model, find a great team to create the hardware and firmware, and select the device architecture that meets the project requirements.

Then comes the work of implementing the app and device – and figuring out how to keep them operational and useful. The potential for revenue generation has increased with subscription pricing models, but the success achieved depends on delivering products that work well and offer unique features that customers are willing to use on an ongoing basis.

Whatever your business plan, the 2022 IoT landscape demands transparency and the ability to meet high customer expectations. Unfortunately, all customers have access to a megaphone, and experience tells us that dissatisfied customers tend to deploy it often. A bug, a device crash or poor customer service – any weakness can be showcased out loud and repeatedly on social media.

If you decide to build an IoT device and want to build customer loyalty and develop a quality product, there are four main things you need to do for a successful launch.

1. Build for the long term

There is no doubt that you will submit bugs. This is not a problem. But what is a problem is not integrating an observability method to monitor performance and fix these bugs.

IoT designers cannot anticipate all the ways end users use their devices. Such use reveals bugs you didn’t expect. Solving these problems requires near-instantaneous access to data and logs. If a product is designed to capture and manage this data, then designers have the opportunity to improve their product before anyone notices. Without access to device data, expect to spend countless engineering cycles trying to recreate a single event or ignore it and wait for others to experience your product in a suboptimal way.

Integrating observability into your devices is critical to the success of your business. Consider other stakeholders who may need critical operational data; marketing may need to know how the number of updates correlates to a reminder email, or the product team may want to know how many users are interacting with a specific feature on a daily basis. Device health and operations insights are critical to your team and all business stakeholders.

If you don’t build in a way to monitor devices and capture critical data, reverse engineering your device to capture post-launch data is much more expensive and painful. Building your product with continuous device observability from the start ensures instant access to the data you need to keep a product running and improving.

2. Use updated development approaches

Traditional approaches often mean product freezes, leaving teams wasting months of development time. Adopting a flexible and observable approach can avoid these delays. Planning and integrating remote monitoring and updating capabilities allows teams to ship a minimum viable product and iterate even when the product is in the hands of the end user.

Using a zero-day update allows you to ship devices with incomplete firmware, then push an update over-the-air to the device when the customer unboxes it. This approach allows you to ship on time, while giving software teams a few crucial extra weeks to create final firmware.

3. Deploy device updates

For device fleets in the tens of thousands or millions, one glitch in a fleet-wide deployment can create a real fiasco. Limiting the impact of unforeseen bugs in release updates is a superior alternative, which can be achieved easily with staged deployments.

If you limit an update to company employees and then to relying parties, it means you can assess potential issues on a small scale without major disruption. After initial deployment, developers should segregate a small percentage of end users for the next stage, monitoring device health and capturing real-time metrics for potential triage and repair.

Another option is time-release intervals. This approach deploys a predetermined interval, during which calling devices receive live payloads. These devices can then be monitored as they register to determine whether to expand the deployment.

With options to limit your builds, there’s no reason to push a fleet-wide update before confirming its success.

4. Take a proactive approach to security

Perhaps because it is a nascent market or because of the speed of innovation, the IoT industry has focused less on security than its software counterpart. The results are serious; sophisticated botnets can use simple devices, such as cameras and hacking devices, to spy on unsuspecting users.

Developers should build a solid security strategy from the start. The basics include securing your updates and cryptographically signing firmware to prevent others from loading firmware onto your devices.

A next step is to use cryptography to secure the link between your device and the cloud, whether it’s Transport Layer Security or Mutual TLS. Building with security in mind not only saves you money, it exponentially reduces the risk of your device being compromised in the future.

Building with an eye on the long term. Build on today’s best practices. Capture insights and learn from deployments. Be proactive. These relatively simple steps can help ensure that your IoT device is set up safely and successfully, and that your customers remain happy and loyal.


About Author

Comments are closed.