09 Sep Product Design: How to Choose the Right Sensor
Sensors are one of life’s unsung heroes: ubiquitous in machinery of all kinds, yet virtually unnoticed unless they light an alarm or malfunction. Think of the warning lights in your car that let you know your tire pressure or brake fluid is low. You may not realize these pressure sensors exist but they’d better work when you need them.
Likewise, in a medical device, a sensor malfunction can quickly escalate to a life-threatening situation. The perfect sensor is simple, reliable, and, in a way, out of mind.
Sounds simple, right? If you’ve chosen the most appropriate sensor for the instrument, then that’s exactly how it should look in the end: Simple. Getting there, however, requires forethought, organization, and planning.
In complex products, sensors may monitor system processes, detect when something is or isn’t present, or keep track of environmental conditions like temperature or humidity.
The process of finding the right sensor inevitably focuses on a few commonalities. You’ll have to consider:
- Environmental conditions
- Longevity – that is, how many cycles the sensor will experience
- How long the product in which the sensor operates will last
You also have to worry about size, sensitivity, and the material. Beyond those generic factors, the process of sensor selection varies from device to device, as you work through the implementation issues specific to the device.
Case Study: The Winding Path to Simple
One of our recent projects — a medical device — illustrates the generic and the device-specific issues we think about when choosing a sensor. We were developing a desktop instrument for use in a sensitive medical fluid pumping application, where a technician operates the instrument to mix a number of fluids prior to patient infusion.
By design, the instrument had a disposable component attached to it, and fluids flowed through that component. The sensor’s job was simple. It only had to answer a single question: Is the disposable there, or not? Answering that question was central to the work flow: If the disposable wasn’t present or was disconnected, the instrument’s hardware and software had to respond appropriately and quickly. Most critically, the instrument had to stop pumping fluids – to avoid exposing the users or the product to risk.
We considered a simple mechanical switch that gets flipped when the disposable sits on it. But that wouldn’t work: The device had to be waterproof. If someone accidentally spilled fluids on it, they might leak in. We moved on.
We considered a waterproof type of membrane switch that had a tiny dome that closes an electrical circuit when it’s pushed down. We rejected that one because it wasn’t reliable enough to keep up with the usage demands of the instrument and required very tight tolerances in the instrument and disposable.
Next, we looked at optical sensors, which work by shining a light at a detector. Optical sensors are inexpensive and reliable, and they’re designed to survive millions of cycles. But they’re sensitive to dirt. Any debris that lands on the face of the detector or the light source can cause problems, and they’re difficult to clean.
Finally, we settled on a magnetic sensor. It’s super-reliable and can perform its job through plastic and non-ferrous metals. It’s inexpensive and easy to install, and we could make it waterproof by simply locating it inside the aluminum device enclosure. The only downside was that we had to install magnets in the disposable. Magnets are relatively inexpensive, but time and effort was required to ensure that the magnets would pass the challenging sterilization process required in this application.
Magnetic sensors aren’t right in every situation. They vary in sensitivity, so we had to be careful about calibration and adjustment during instrument assembly. Since we were only making a few hundred of these instruments – and not, say, hundreds of thousands – the labor time and cost of assembling and adjusting the sensors wasn’t outrageous.
The Value of Experience
Every product is unique, but that case shows how the process of selecting a sensor should work. It’s best to think about sensor selection early in the process of product design, and to think broadly. You need to consider implementation, procurement, assembly, and cost. And you’ll want to think about the risk involved in choosing a sensor – including technical and development risks, and those related to schedule and costs.
Our experience over the years has taught us that something as humble as a sensor ripples through the design of a product in surprising ways, but we can anticipate those ripples and, at the end of the design process, find the simple solution. Want to talk more deeply about sensor selection, or another technical problem you’re facing? Just reach out.