Bargain-hunting engineers say sensors still too costly

Some sensors are so ridiculously cheap it almost seems counterintuitive that engineers cite budget constraints as a challenge.
Some sensors are so ridiculously cheap it almost seems counterintuitive that engineers cite budget constraints as a challenge. (Flickr)

The market for sensors is big. And it’s getting bigger, with design engineers packing more sensors into everything from automobiles to smartphones and applying them in new applications like smart toothbrushes and cat litter boxes.

As sensors proliferate, concerns over cost continue to preoccupy engineers, even though sensors frequently make up only a small fraction of the total bill of materials (BOM) for a product.

A case in point: in a 2019 teardown of the Samsung Galaxy S10 5G SM-G977N, TechInsights estimated the cost of the category “sensors” at $2.50, or roughly 0.5% of the total hardware cost of $490.06.

Sponsored by Digi-Key

Industrial Ethernet Solutions from ADI Chronous™ Available Now from Digi-Key

ADI’s Chronous portfolio of edge-to-enterprise Industrial Ethernet connectivity solutions is designed to accelerate your path to Industry 4.0. The compatible and interoperable Industrial connectivity products enable best-in-class performance.

In fact, some sensors are now so ridiculously cheap it seems almost counterintuitive that engineers would cite budget constraints as a challenge.

Yet, in a recent survey by FierceElectronics, more than one-third (37%) of engineer-respondents who design with sensors cited “meeting budget constraints” as a top challenge, only superseded by a nemesis all too familiar to engineers—sensor integration.

The cost of sensors is a concern for product engineers.
The cost of sensors is a concern for product engineers.
Source: Sensors Expo Survey Results

Every Penny Counts

“For some high-volume consumer products, pennies do matter, and the actual sensor BOM cost is critical, says Matt Liberty, CEO, JetPerch LLC. Liberty knows a thing or two about keeping hardware costs in check—he is the creator of Joulescope, a product described as “the most affordable and easy-to-use precision DC energy analyzer.”

On the other hand, Liberty points out that for many other products, the cost of the BOM itself may be small compared to other factors, including the engineering development time, the increased product size, and the cost of the components required for computation. But even then, cost is a consideration. “My mantra is ‘When in doubt, leave it out.’ Excluded features are free!” says Liberty.

Michael Anderson, Systems Architect, Huntington Ingalls Industries, says that the total cost of even budget-friendly sensors can really begin to skyrocket when products go into mass production. 

“If a sensor is $5, and you need to produce 20 million devices (e.g., a cell phone), then that's $100M for the sensor array. So, for hobbyists, the cost is inconsequential.  But, for mass production, these sorts of things tend to add up” he stresses.

While at that point the cost of the BOM for production starts getting in the way, Anderson says that engineers also have to factor the circuit into their design, consider the possibility of licensing the IP to put it into an SoC or duplicate it in an FPGA, etc. 

“So, as easy as it seems, there are complications that many engineers are simply not qualified to resolve because they're more financial than technical,” says Anderson.

Pressure to Keep Hardware Costs Low

As the percentage of software in a product increases, and the value and differentiation that the software brings to the product increases, there is further pressure to keep hardware costs low.

“Hardware companies making money today are doing low volume and complex, integrated systems—things like access control, scooters, mesh network sensors, farm animal trackers, etc. In these products, the hardware isn't special. It's a sensor that reports a value back to a central server,” says Tyler Hoffman, a co-founder of Memfault and an embedded engineer with experience developing consumer products like the Pebble Smartwatch.

He stresses that the value of these products comes not from the hardware, but rather from the centralized cloud infrastructure, the API's, mobiles apps and websites, as well as the storage and processing of data—all of which requires a cash infusion on a recurring basis,

“So,” he says, “In many cases the reality is that the hardware must be made as cheaply as possible.”

To be sure, efforts to develop lower-cost sensors continue apace. For example, at this year’s Consumer Electronics Show, sensor maker Velodyne LiDAR introduced the VelaBit, a compact, automotive grade LiDAR solution for autonomous vehicles. At a price tag of $100, it’s significantly less expensive than the company’s other LiDAR offerings, like the $500 ULTRA Puck.

But as is often the case, that lower cost sensor comes with a compromise in performance. The VelaBit’s range is 100 meters, exactly half that of the ULTRA Puck.

And swapping cost for performance is just the kind of trade-off that some design engineers may not be willing to make.

Suggested Articles

Artificial Intelligence can be used to speed up ingestion and processing of data in supercomputers, according to Nvidia

Advances in CMOS sensors are making the technology a key source of data in applications ranging from smart phones to self-driving cars.

Vehicles to ventilators: Scaling from a small number to many thousands of ventilators per week is a challenge, along with finding the right parts