Problem solve Get help with specific problems with your technologies, process and projects.

The unseen killer of IoT devices

IoT is many things these days, but uncomplicated is not one of them. At every turn in the road, there are obstacles such as cyber threats, wireless signal interference and more. While many IoT device makers spend their time working to address these issues, they often fail to notice an equally critical obstacle lurking in the shadows.

It’s an obstacle that goes unnoticed because it’s internal to the IoT device rather than external, such as an attack from a cybercriminal. But make no mistake: Any product maker that doesn’t face it head on runs the risk of having their product fail once it’s in the end users hands. It’s a mistake that can impact a company’s bottom line and brand, while also making it vulnerable to costly recalls.

What is this obstacle, this unseen killer of IoT devices? It stems from using measurement instrumentation with insufficient bandwidth to test IoT devices. Essentially, it boils down to one simple fact: you can’t see what you can’t measure.

Suppose a security camera were set up to capture an image every few minutes. If something happened when the image was taken, it would be readily captured and documented. But what about those precious seconds and minutes between image captures? If something out of the ordinary took place in this span of time, it would go completely unseen by the viewer.

Measurement instrumentation works in much the same way. When it doesn’t have sufficient bandwidth, the user risks missing transient digital events. This might seem like a trivial thing, but nothing could be farther from the truth. One testing error or missed transient could cause an IoT device to fail prematurely.

Just imagine that a chain of 10 hospitals installed a total of 20,000 smart dispensers containing hand sanitizer, and they all started to go dead long before their batteries’ service life was up. That’s 20,000 batteries that would need to be replaced. Assuming one technician could replace 100 batteries a day, it would take 200 days — or approximately 6.5 months — to replace all 20,000 batteries. That doesn’t even begin to take into consideration the cost of the new batteries, any associated disposal fees, and the hit to the product maker’s reputation and relationship with the hospital when the smart dispenser product didn’t perform as expected.

Source: Shutterstock.

The issue here is that all of this came about not because the IoT product maker didn’t do the right thing. Its design and test engineers came up with a solid design. They optimized it to deliver the best performance. And, they fully tested it to ensure its functionality would be as expected in the real world. But the problem is that the designers picked a measurement instrument to conduct that testing that had insufficient bandwidth.

That single critical mistake meant that even though the device was thoroughly tested, a problem went undetected. As it turned out, the smart dispensers were programmed to routinely notify the hospital’s facilities team of their operation and any maintenance or refill that was required. That notification was taking place too often and draining the battery far quicker than anticipated. Had the engineers employed a measurement instrument with sufficient bandwidth, such as one that could perform fast measurements continuously to precisely capture transient current waveforms and ensure accurate current consumption characterization, this problem would have been detected.

The bottom line here is clear: if you can’t measure the IoT device accurately, you run the risk of missing problems. It’s a painful lesson for any company to learn the hard way, but it can be avoided. It’s as simple as picking test and measurement equipment with sufficient bandwidth.

All IoT Agenda network contributors are responsible for the content and accuracy of their posts. Opinions are of the writers and do not necessarily convey the thoughts of IoT Agenda.

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.