Friday, September 12, 2014

The IoT - Odds and Ends


In this week's blog post I offer up a selection of thoughts and observations which individually aren't large enough to encompass a discussion with a full editorial.

Dark Horses

Oft expectation fails and most oft there
Where most it promises, and oft it hits
Where hope is coldest and despair most fits. - 
William Shakespeare, "All's Well That Ends Well

Apple's product announcements on Tuesday created a media frenzy the likes of which hasn't been seen for years. Granted, the iPhone 6 and 6s, with larger screens and greater LTE bandwidth for potential worldwide service coverage are attractively detailed offerings that offer decidedly interesting feature enhancements. The Apple Watch is also a well designed device that will certainly appeal to wearable gadget aficionados. It's extra rich feature set was an insightful calculation intended to rock competitors back on their heels and will undoubtedly throw earlier entries from Samsung, Sony and others on the defensive for a time as they scramble to play catch-up.

Yet at the risk of sounding churlish, I am much less sanguine than the mainstream media on the long term effects of Apple's new products on the smartphone market. The segment appears to be reaching saturation this year as unit shipment growth seems to be fast closing in on an asymptote, with sales likely to be flat or possibly in slight decline for 2015. I personally do not see Apple's iPhone 6 line re-energizing the smartphone market and vaulting it back into the levels of growth seen at the beginning of the decade. The new iPhone releases, though, should help Apple reverse its declining market share trend and begin to take back share from its rivals, who in turn will need to find lower price points and feature enhancements for their own future releases to stem the tide. 

As a side note, there may be something meaningful in the developing trend towards larger smartphone screen sizes. First pioneered by Apple with the original iPhone blowing Blackberry and Nokia out of the water, trumped afterwards by Samsung and now with Apple following suit, this raises the possibility that smartphones and tablets may be on a collision course that amounts to convergence (more on that later.)

But the more egregious lapse in judgement by the mainstream media - the one which lends justification to the notion that hooking up the average reporter to an EEG would generate a perfectly flat readout - is their assessment of the Apple Watch. Their error was in accepting the portrayal of this product as an IoT device. Last week's blog post took pains to separate the wheat from the chaff in order to delineate the qualities a product needed to have in order to deserve such an appellation. Viewed in the light or those requirements, the true nature of the Apple Watch becomes apparent - it is an astutely designed iPhone accessory. Its functionality has nothing to do with the internet; thus, it cannot be legitimately called an IoT offering.

If we consider the Apple Watch in this context, we can more objectively assess its impact on Apple's fortunes. As an apparel choice, some might find its shape aesthetically appealing. Others won't care and will only be interested in its features and functions. But the truth of the matter is that many people who still wear watches DO care about aesthetics - and this is where Apple has grossly failed in its design conception. Apple has in fact repeated the mistake made by Google with Glass - they have created a product whose features and benefits are encapsulated in a form factor that has significance to an individual consumer's sense of fashion and their outward appearance in public venues. This is bound to negatively impinge on the TAM for the Apple Watch.

Apple has also unwittingly provided Samsung, Sony, Motorola and any other wearable competitor the avenue for creating a rival product to displace the "iWatch." It wouldn't surprise me in the least to hear of Samsung signing co-development agreements with Patek Philippe, Cartier, Rolex or any of another half dozen elite watch companies and releasing a series of stunningly beautiful and amazingly functional smart watches in the next 12-18 months. (This would still NOT be the "killer app" of the IoT, however - more on that later as well.)

Though it may sound like I'm picking on Apple, I'm actually trying to make an altogether different point. One can't help but notice that consumer electronics giants the world over have failed miserably in creating a wearable 'killer app' over the last several years. None of their products in the IoT category have achieved that status, and as noted last week, one in particular - Google Glass - has actually damaged the company's reputation and public image.

From my point of view, everyone observing the consumer electronics giants for developments and direction is looking in the wrong place. Their very success in PMPs and mobile phones may actually be predisposing them to fail in the IoT.

A dark horse which had never been thought of, and which the careless St. James had never even observed in the list, rushed past the grandstand in sweeping triumph. - Benjamin Disraeli, "The Young Duke"


From last week's editorial, we can discern a very distinct pattern in all the non-consumer applications that are touted as IoT solutions. A network of sensors with variable levels of native functionality are tied thru a wireless spectrum to a controller, which in turn is slaved to some sort of user-managed server, either thru a LAN or wireless band. The server can be a desktop, laptop or smartphone. There is, in fact, a variation of this development trend underway in the wearables world, wherein the wearable is actually an integrated system of sensors with their controller, again slaved to a server (a smartphone in this case.)

When viewed from this perspective, one can deduce that an instinctive appreciation of genuine IoT applications in the non-consumer part of the market lies not with consumer electronics firms, but with companies that build sensors and MCUs. Some will recoil in astonishment that I could advance such a claim. After all, the major MCU and sensor firms tend to have miserable margins and command none of the awe and admiration of IP-rich companies such as Broadcom, Marvell and other 1st tier SoC vendors. 

Yet it is acutely evident that the IoT market, still in its infancy, is evolving along lines that suggest it will inexorably gravitate towards these sorts of enterprises as suppliers. Indeed, companies such as NXP, TI, Freescale, Avago and Microsemi are not only particularly well suited for evolving into the primary developers of IoT peripherals thru their products, customer portfolios and institutional histories, but may find themselves inevitably and even unavoidably propelled into such roles. 

Consider the evolution of the chip industry from discretes to SSI (TTL, 8b CPUs, PALs and GALs) to MSI (controllers, PHYs, 16b CPUs, CPLDs and such) up to system-level LSI (SoCs, top end FPGAs & ASICs, 64b CPUs and so forth.) This growth in technical mastery over hardware, software and toolchains was paralleled by a transfer of value from systems down to chips - value which is now transitioning to IP vendors and, eventually, to the foundries themselves, as discussed in earlier articles. It is an inescapable reality that this same transfer of value will be imposed on sensor & MCU companies by their system level customers, demanding ever greater levels of hardware integration and software functionality at the same or declining levels of TCO (total cost of ownership.)

What I predict we will see over time is tablets and smartphones converging into a genuine Personal Processor, with everything else - including screens, keyboards, headsets, cameras and other sensors - treated as wirelessly connected accessories and extensions. The PP and screens will likely continue to be properties of the consumer electronics titans, while all the other peripherals would devolve to integrated sensor and MCU companies that had extended themselves into the subsystem level.

Brave Achilles

The apple cannot be stuck back on the Tree of Knowledge; once we begin to see, we are doomed and challenged to seek the strength to see more, not less. - Arthur Miller

Some time before 1000 B.C., the greeks were a very different people from those that we know from much of the classical literature. They were a feudal people, with warlords in semi-autonomous city-states dominating the local peasantry. We refer to them today as Mycenaeans, whereas the Greeks of the classical period generally called them Achaeans.

Many myths were passed down from those days as oral traditions. The most powerful of those stories include Homer's Iliad, which recounts a fanciful version of major events during the Trojan War. Achilles is a central character of the tale. He was the strongest and fiercest of the Greeks, nemesis to Hector, Hero of Troy and first son of Priam, its king.

Of semi-divine descent, Achilles was prophesied from birth to experience a heroic but ultimately tragic destiny. His mother, the sea nymph Thetis, sought to thwart this augury by bathing the infant Achilles in the river Styx, the fabled river of forgetfulness which separated the land of the living from that of the dead. It was thought that the waters of the Styx would convey invulnerability on the child, and this proved to be true, as so many ill-fated Trojan warriors discovered to their dismay. However, while bathing her infant, Thetis held onto her son's ankle, leaving him vulnerable in that part of his physique. From this oversight, Achilles ultimately met his end scaling the walls of Troy when he was struck on that assailable heel by a poisoned arrow shot by the reviled Paris, another son of Priam. It is from this tale that the phrase "Achilles heel" has been passed down to us - an unsuspected yet fatal weakness in what is seemingly an otherwise unbeatable foe or irresistible force. 

Some of you are likely saying to yourselves at this moment "Oh, brother, not this baloney again - Hey, Baldie! How about getting to the point? Some of us don't have all day." I promise you that there is a point to my recounting this ancient legend, and I'll be brief.

Below is a snapshot of the latest measure of the Consumer Research Bureau composite price index. This is a very powerful and surprisingly overlooked indicator. It measures pricing activity across 23 raw materials categories for industrial, textile and agricultural markets. From this, one can deduce the health of the global economy and get a feel for price level changes and, hence, the waxing or waning of inflationary forces. This data is, in fact, a far superior inflation indicator than the CPI, which is heavily compromised by the federal government's grossly warped measuring criteria that are designed to significantly under-report inflation.

There are multiple sub-indexes that can be sourced from this data, with all of it available at The sub-indexes tell much the same story - one that is utterly at odds with the daily prattle from the MSM talking heads on cable and in the print media and their 6 year old 'continuing recovery' gibberish.

As can be seen from the chart, commodity prices have been mostly stagnant since 2011. This year's first half run-up has been fading since early June, strongly suggesting that the raw materials stocking for consumer products build-outs before the Christmas and New Years holiday season dropped off earlier than usual - an indication that manufacturers and retailers are likely approaching the Q4 shopping season with caution. 

What this graph plainly illustrates, then, is a global economy that has been stagnating for the last three years. Such stagnation creates insecurity at the corporate, small business and individual wager earner level. Furthermore, in the United States - the premier consumer economy in the world - inflation-adjusted wages, median household incomes and median household wealth have been in decline for multiple decades.

One could write a book about such issues, replete with a cornucopia of tables, graphs and charts. For the purposes of this editorial, we can confine this portion of the discussion to some evidence supporting the above assertion - evidence that is available at the following links:

Here is where we find the Achilles Heel of the IoT effort - consumer disposable income. The combination of a soporific economy, job insecurity, depressed household wealth and wages that aren't keeping up with inflation all serve to constrain how much extra cash people are willing to shell out for non-essentials. 

Most wearables are priced between $50 and $400 retail, depending on the device. Such an expense is not a serious issue for people in the top 10% income bracket, but does become a consideration for the other 90% of consumers. And it is in that 90% where the potential customer pool for a "killer app" resides.

This doesn't mean that no one will buy wearables except the wealthy and well off, but it does put a significant burden on wearables developers to create products with truly compelling value propositions. Put another way, it doesn't indicate that a 'killer app' is simply beyond the reach of any company, but it places a great deal of additional pressure on developers to create products with a combination of utility and user enjoyment at a given price point that is truly remarkable. What we have seen so far in the IoT market is interesting, but it is nevertheless evident that the growth market for High Tech which will take over from the flagging smartphone and tablet segments has not yet been discovered.