In my previous blog post, I mentioned the Linley Group’s 2014 Mobile Conference. In past years, the focus has been on the mobile processors (and other chips) found in smartphones, but this year another type of mobile product gained significant attention: wearables, such as the Fitbit Flex wristband and Google Glass. There was wide agreement that wearables will become an important product category, but virtually no consensus on the kind of product we’ll see over the next ten years.
Wearables are today where the smartphone market was in 2003; lots of exciting stuff is going on, but the category-defining product had not yet been introduced. For example, some wearables today are designed as peripherals to a smartphone, simplifying both their communication needs and the kind of mobile processor required, while others have sophisticated processors, sensors, and displays. No one really knows which design direction will dominate in the future.
Pankaj Kedia of Qualcomm observed that wearables are really a subset of the Internet of Things (IoT). At their simplest, wearables consist of one or more sensors connected to “just enough” signal processing silicon and a method of communicating that processed data, such as Bluetooth or Wi-Fi. I agree, and expect that on a volume basis, this kind of very simple, very low cost product will predominate. In ten years, every shoe could easily contain a built-in pedometer. Why not, if it adds less than 1% to the cost of the shoe, and everyone’s smartphone is running a health monitoring app which can accept this data?
But as we look at complex wearables, such as a smartwatch, the future is harder to predict. At one level, it’s not clear if a smartwatch is meant to replace a smartphone, or simply be a peripheral for it. The answer will determine the kind of computing and connectivity semiconductors that will be required. And a vendor’s ability to correctly predict which kind of product to build will determine its marketplace success. But at this stage, conference panelists admitted, it’s not really clear if a smartwatch is intended to replace a regular watch. (Personally, I wear a watch given to me by my father; the Samsung Galaxy Gear smartwatch is pretty slick, but it’s not going to push it off my wrist any time soon.) Figuring out how consumer electronics are going to intersect with the fashion world, as well as the public’s general concerns about these products from a privacy perspective, is much harder than comparing technical specs to predict which System on Chip (SoC) device is going to sell the best.
Chris Anderson, Wired’s Editor in Chief, called wearables “the peace dividend of the smartphone wars”. The incredible competition to produce ever better handsets at ever lower prices has brought the price of many components (particularly sensors) down to commodity price points – so why not include them? For example, the price of a three-axis accelerometer has dropped from $4.50 to 20 cents. Kurt Shuler of Arteris outlined some very compelling use cases for wearables in the medical and safety fields, such as glucose monitors for diabetics and temperature sensing clothing for fire fighters. But there was general agreement with his observation that in the broader consumer market the “killer app” has yet to emerge.
It’s always interesting to hear tech savants speculate about the future. We all know that somewhere out there, some startup will combine these components in an unexpected way, create a product no one’s yet conceived, but which we’ll all clamour to own. And this will be possible because the inventors had the value of their R&D protected by a strong patent system. Furthermore, there’s a good chance that they’ll make use of licensed IP blocks, such as ARM processor cores or DSP cores. And should this startup stumble and fail, their venture backers know that their investment is at least partially protected by their ability to sell or license any patents filed by the company. A robust environment for protecting intellectual property is essential at every stage in the process of innovation.