The future of the sensor market

I’ve written previously about how declining sensor costs are a key facet of IoT. Along with cheaper compute costs, better (and also cheaper) cloud integration and improved mobile broadband, sensors round out the keys not just to a networked physical world but a world of intelligent machines and machine learning.

So all is hunky dory if you’re a sensor manufacturer? Sort of. On the one hand your unit growth is ticking along at over 11 percent a year, according to one estimate. Just look at Samsung. Accelerometers and gyroscopes weren’t enough for their smart phones. They’ve thrown in pressure and humidity sensors.

But on the other hand your average selling price is forecast to fall by 5 percent not just this year, but every year for the next five years. Let’s, um, hope your input costs decline.

The trend is driven by intense competition as well as basic commodification of different sensor types. IP appears difficult to lock up in the market. Throw in the reality that many IoT applications require rock bottom pricing and that puts pressure on sensor manufacturers not just to lower prices but also to undercut competitors in the hopes of locking up the market for themselves, Amazon style. The giddiness for consumer IoT has infected sensor makers and no one wants to tell shareholders that they can’t sell sensors to consumer IoT companies when that market is growing.

If you’re thinking consolidation, you’re partially right. There isn’t enough room in this tight market for all the players, but more of what’s going on right now is that companies are looking to bulk up their sensor portfolios through acquisitions.

There have been no shortage of acquisitions. Sensor maker InvenSense’s purchase of Movea and Trusted Positioning has helped the company with indoor and outdoor navigation as well as motion sensing. In another deal TE Connectivity bought Measurement Specialties for its sensor portfolios of everything from vibration to ultrasonics sensors. More recently, Sony snapped up SoftKinetic for its image sensor technology.

A couple of things are happening. First is that some sensors like gyroscopes are commodified, while others like complex indoor motion sensors are still in the viable IP category. Second, if you’re a company working with product manufacturers you need to be able to offer a suite of sensors.

And so the name of the game going forward is being able to offer a strong portfolio of sensors for customers as a hedge against downward pricing and margin pressure in many segments.

The other market trend on the horizon is integrated modules. Intel’s big consumer IoT announcement at CES of its Currie chip heralded the company’s effort to win over consumer IoT. The module includes not just a 32-bit Intel system on a chip, but also a DSP sensor hub and a 6-axis combo sensor with an accelerometer and gyroscope.

For some IoT product developers there’ll be added value in having sensors packaged with processors and even communications chips (Currie has Bluetooth LE built in). This is all the more reason why I think it’s possible we could see more traditional microcontroller makers going after sensors makers.

So it’s the best of times and the worst of times. If you’ve got a great new sensor with some valuable IP, excellent. If you’re trying to compete on price with commodified sensors, not so much. What’s clear is that for many manufactures in markets ranging from processors to connectivity, making a sensor company acquisition is worth consideration.