Jan 9, 2017

38 min read

A Very Producty #CES2017

Observing and learning from the show floor 👀🎓👍

CES 2017 was about execution and iteration, and there was a lot of those on display. While this year’s exhibition had all the usual fanfare of big company “launches”, it took a little more digging below the surface to see that the industry is making a lot of progress relative to announcements over the past couple of years.

If you come to CES expecting to see all the new innovations or to place orders for products hoping to beat you home, then it is easy to become a bit cynical or even worse to become jaded over many years. With my own history of 26 CES shows behind me, I can safely say this was one of lots of new, cool, and useful but not any “firsts”. With that said it is also worth noting that even the firsts of the past (some are detailed at the end of this post) were not actually first at CES nor received with massive excitement (usually it was more puzzlement) when they were announced.

The theme for me for this year is “Year of Product Execution”. By this I mean that from my perspective what I saw was “another turn of the crank” or maturing for many categories. Products are crossing from “invention” or “technology” to product or category. The products that have been out for a couple of years are consolidating, with fewer makers that have more complete offerings. This doesn’t apply to everything as I’ll note below, but across the board in the innovation categories it is clear another year of progress has been made.

This was a well-run CES. The final attendance numbers are not out yet, but the last day press release said 175K which is below last year’s 178K. The only painful crowds were at Samsung and LG, and also the CES run innovation pavilion (which just has way too many winners now, imho). There’s a good chance the shifted calendar caused a bit of unintended thinning (things are back to Wed-Fri for 2018).

My goal is not to find the best or the coolest but to focus on trends over time and to highlight those trends with examples. The report is organized with a couple of sections up front on strategy and industry observations and then on my own view of categories:

CES Is Not For Shopping!
Thoughts For Product Managers
My Tops
Voice and Cast
Auto
IoT
Mobile OS/Chip Devices
TV
Drones
VR/AR
PC
Health
Imaging
More…
Bonus: CES History

Some author notes. This is just my report of what I saw and thought at the show and does not reflect any companies I work with directly or indirectly, or might work with (past or present), and certainly not the views of a16z, Box, or anyone else. While obvious, as one person I can’t possibly “cover the show” and I miss lots no matter how hard I try. Please just do a web search for more information on any products mentioned here as my intent is not to do a full description or review — there are many great tech articles detailing each product. Omissions or errors are just my own limitations. Photos are my own unless indicated.

CES Is Not For Shopping!

After seeing so many tweets skeptical of CES or reporters bemoaning the show, it seemed worth a quick recap for why companies attend the show (btw, TV reporters still tend to offer rather breathless coverage compared to twitter or online). Skip this section if you’re just excited about new stuff.

To begin with the show is not for shopping or buying, even though so much of what you read often expresses some frustration with availability or lack of information on pricing/availability. It is not a retail environment. It is also not for product support, though it is incredible to see how much of this goes on. The popular direct to consumer products always seem to have a lot of “when I do this” or “how do I do that” going on at the booths. For example, I saw a Nikon rep spend a good 20 minutes on how to configure Auto Focus on the high end DSLR which reminded me of staffing the booth answering questions about bullets and numbering.

While there are many reasons why a company might attend the show, here are some main bigco reasons:

Show off innovation to press or mark territory for specific industry press. A company looking to let its own industry know it has an offering in a new area or to stake a claim on a new area often uses CES to announce something. Such announcements are often staged outside the main floor or behind glass cases. These might be whole products or just specific technologies. Up until this year, most every OLED TV (excepting LG) was exactly that and though only LG was in market, every maker was showing off OLED. This year everyone has products in the space.
Validate complex web of partnerships. Most every big company is working to create or support an ecosystem around its major products — these are often rather complex webs of relationships that require both parties to show off the relationship. It is also why the big company booths often have clusters of small stations showing off third party add-ons or add-ins for the platform or products. These partners can’t often support their own booths or travel to the show, so these partnerships are subsidized and have the primary purpose of showing everyone else the depth of support. Qualcomm is a great example of an exhibit that primarily exists to show off the broad range of industry support is maintains.
Bootstrap new partners or investments. Of course many companies do choose to launch new products at the show or let the industry know about a newly forged set of partnerships. When a company kicks off a whole new business or especially when two big companies get together, you can usually see some exhibit space (as well as offline briefings and announcements prior to the show) highlighting the partnership. In many ways, the entirety of the Auto pavilion has become as much about which company is working with which other company as it is to show off cars.
Vast array of channel middle players doing business. The original/primary goal of the show is rooted in the complexities of supply chain and product distribution and support, and those players are still an integral part of the gathering. While we like to think everything just magically emerges from China and shows up on Amazon, there is a massive amount of supply chain, wholesale and retail distribution, as well a post manufacture sales and support that many products require. Head over to the A/V section and you can hear installers talking about programming remotes or wall space required. Listen in on any of the mobile phone case booths and before long you will start to hear about volume commitments and delivery timelines. In many ways this is the un-glamorous part of CES that keeps the lights on!
Compete. Perhaps the most exciting is that the show is also about both gathering and putting forward competitive understanding of the various categories or markets. There’s nothing more fun than heading on over to a competitor’s booth and seeing what can be gleaned from those staffing the booth or simply getting to see the product before it is actually available. I loved listening to co-workers of one of the WiFi router companies talk during a demo of another router. They were talking about the use of USB-C power cables (a real plus that I also noticed, more later).
Startups and Geographies. I wanted to call this out because so much of the show is now working hard to bring attention to these two areas specifically. Geographies have always had a special place at CES, with dedicated space for “countries” to support or subsidize booths. Believe it or not, the only place to see a Samsung or LG years ago would have been at the Korea booth! Since most countries have government offices focused on fostering growth, innovation, tech, and/or startups these areas show off their work. Startups are relatively new and, at least for me, a tough fit in the show. I have to think it is incredibly expensive (time and money) to attend and the chances of landing big sales or that next round of funding amongst all that is going on might not be so great. I don’t want to be negative, especially because everyone starts somewhere (Microsoft once had a single table at COMDEX!) but times have changed substantially and so too has the approach of go to market. This year featured a good deal of space devoted to Indiegogo and Kickstarter projects before they are even listed on the crowd-funding site. It will be interesting to see how this turns out.

Not every company exhibits at CES (though every company has people here, and big companies might have 100's!). Relative to the industry the show is in a weird stage where several companies that dominate the consumer space or provide infrastructure across the space are not directly present (Apple, Google, Amazon, Facebook, etc.) Exhibiting costs millions of dollars and takes a lot of time and comes with high expectations, and right now these companies basically are everywhere without having to do all that work. It is becoming increasingly weird though. As PCs rose (and COMDEX declined) it is why Microsoft increased its presence at CES (Microsoft no longer has a large formal presence, nor do any of the big PC makers).

Thoughts For Product Managers

If you’re releasing a product into the environment covered by CES then there are some broader strategic observations from the show worth sharing. This is hardly everyone, for the most part enterprise software is way out of place at the show. The industry continues to shift and watching the changes over time can help to solidify these strategic changes and help to focus product development efforts.

Software is the limiting factor, 100% of the time. While the attention and focus of the show, and all of the biggest exhibits, primarily go to market with hardware, it is software that is constraining everyone. It is hard to find an example of a product where the software is the biggest gating factor between “me too” and “breakthrough” or even where software can make the difference between “useful” and “usable” or even “desirable”. The most sustainably differentiated products are the ones with fantastic software. In the short term, it is entirely possible that a bigger company with poor software, or a product that is amazing hardware with not so amazing software will succeed, but it seems at this point the differentiation is not sustainable. As a corollary, a product that is not differentiated by software, arguably is not differentiated at all.
Hardware can (and will) always be thinner, lighter, cheaper, better designed, and get longer battery life, so move on. This has both always been true and never more true than today. CBS Sunday did an awesome story on VR/AR where they showed their own CES stories over the years showing VR headsets that is definitely worth watching as a reminder of how much things can change. The strategic view of this is not to worry though timing matters of course and more importantly not to be critical of a newly created product because of these attributes.
At planetary scale, quality matters more than ever. When serving tech enthusiasts as has been the case for the first years of most innovations and for the first 50 years of CES (that is looking at the penetration of any given device relative to the population), it is clear that all that was lacking in past products relative to stability, fragility, reliability, and more will be intolerable in the next wave. The old jokes about VCRs always blinking 12:00 or no one able to set up an answering machine pale seem so incredibly trivial when you’re talking about markets of a few million units, even though at the time those were huge. The markets of the 2000's and PCs getting to 1B installed and used devices were supported by a literal army of professional support people (plus para professionals fresh off holiday support duty). IoT, VR, AR, Wearables, Health, and more will be severely constrained if these do not become order of magnitude easier. Quality means more than ease of use. It means not having visible attributes like firmware, reboots, and errors that can’t be fixed by a power cycle. These might be real technical “musts” but it is just as true that these are real consumer “must nots”.
At planetary scale, the whole matters more than parts. Every product category wants to use their own economies of scale (i.e. If you make components than you want to focus on using your components) or their own expertise (i.e. If you make software then you want to outsource the yucky, low-margin hardware business). The breakthrough products will continue, more than ever, to be the ones understand that the whole and the seams between the breadth of partners required for a product, are what matter more than any one choice. Additionally, the understanding the seams that truly matter to customers and your own differentiation will be super critical. This does not mean everyone needs to design their own SoC or Bluetooth stack, but it means depending on what you’re doing you might need to.
(Client) Platform Duopoly. The platform duopoly of Android and iOS was broadly on display at the show and regardless of personal preference or “share numbers” there’s subtlety to making platform decisions. On the floor it is super common to see anything that involves deeper access to hardware or more expansive App Store rules to be on Android, but recognize this will likely remain Android-only and so ask whether or not there are some fundamental assumptions to revisit before going down this path. Targeting iOS was clear across everything that had a luxury or upper economic feel, and conversely for emerging markets or just outside the major western markets, the booths typically showed Android. We’ve talked about this a lot in podcasts but if you’re targeting both platforms then it is really important to do both very well and not to think of first/second class or priority. As a side note, there’s a similar situation forming on the cloud services side but it is still (t00) early and is much more of an AWS dominating, Microsoft for enterprise, and Google compete with several others.
Software is tooling constrained. A huge part of the software limitation is that world of making software is constrained by tooling. Since there is no dominant end-to-end platform (i.e. app+service) the tooling suffers from this same seam of front end/back end. It is very tricky for a team of a couple of people to build a great app and a great service together, while also using new techniques such as AI/ML. The tooling for the advanced web (browser+service) never really caught up to the needs and with software in so many things this is a real limitation now that everything needs apps+service (and maybe browser too). Every engineer working on the floor tells you about their execution challenges across app and service.
First Party versus Ecosystem. This is a year where all the major platform players (who are not at CES) seem to “compete” with many of those at the show. From this “competition” a lot of interest gets generated but behind the scenes nothing causes more consternation than these “strategic” efforts. Google has Pixel and Chromecast, but wants everyone to make devices running Android/ChromeOS and have cast support. Apple has Healthkit and Homekit (and many first party apps from News to Keynote to Maps) and wants all those vendors to build on top of these APIs. Amazon sells Echo and supports Alexa as an SDK for other similar devices. Microsoft has Surface and supports PC makers. These relations are tension-filled between giants like Samsung or LG and Google for example, and the competitive threat looms constantly. In fact, right now these first-party competitions mostly seemed poised to be the least optimal for all parties—the premium devices come from first party but represent a fraction of the market, while the broad customer base struggles with the broadly distributed, low-margin solutions. It is worth keeping in mind that in the broad history of competition from first-party, it is almost never the case that the first-party product comes to dominate their own ecosystem. Microsoft Windows and Office is generally the exception to this rule, but it came with a platform shift that the existing App ecosystem refused to embrace.

My Tops

I have never done a top picks or even suggest “best of” but want to try something new this year because two products jumped out at me as doing great work symbolizing the theme of execution. There were many wonderful products and so don’t read too much into this.

Quilla by QuirkLogic. Quilla is a “connected e-writer”. What that means is that it is a 42" Eink display that you use as a “white board”. Eink (e-ink generically) has been around for a long time (rooted in electronic paper ideas from MIT Media Lab in the 80's!). For the past 15+ years the company, eInk, has been executing and working to improve all aspects of the base technology. Of course we all use it in the Kindle and have seen the improvements there. After much iteration and attempts, this year brings the large format display with touch and a connected cloud service from Canadian startup QuirkLogic.

The 20lb screen can be hung on a wall or on a wheeled easel. It can run a full (24h) day on the internal battery and is connected by wired or wireless networking. It has a marker sized pen that allows you to draw and erase, along with controlling the very “app-like” software platform. The software has an expected notebook and drawing surface metaphor. You can of course import PDFs or share the board contents easily. You can access the notes from a mobile app as well.

The connected part is that multiple displays can be treated like one giant board OR you can have boards in multiple locations and they just stay in sync. Think about that for a sec — you draw in one room and it is mirrored in another. Super cool.

Now immediately you’re thinking “G-Hangouts does screen sharing” or “this can be done with XYZ app”. While that is true, think about the fragility of all of these solutions. How much effort goes into setting things up, switching the monitor from slides to the note taking app, and so on. I’ve never seen any of this routinely working.

You might also be thinking of the super cool Microsoft Surface Hub, which is a fantastic screen and packaging for this conference room scenario. The Quilla has all the benefits of a “mobile based” solution when it comes to instant on, power management, security, and manageability. We have to be honest and admit that the reliability of a “kiosk” PC is not generally enough for these unattended scenarios. There’s also benefit from notes being on a second screen from the projected images, when you think of meeting flow like brainstorming and the like.

This is a “point” solution that rises above paper and below “just use a PC or Mac”. There’s lots to consider as it expands but the completeness or execution and reliability really caught my eye.

Ring Floodlight Cam by Ring. Floodlight Cam is the latest product+service from Ring, a California based startup that makes the much liked and copied Ring Doorbell. Floodlight Cam a motion-activated HD security camera with built-in floodlights, a siren alarm and two-way talk.

Again this is a story of execution. Ring has been on its mission to reduce crime since 2012. I’ve used each of the first two generations of the bell and it totally changes how you think about the basics of home security and safety. By merging the routine of a doorbell with motion, camera, microphone things really change. From wherever you are you can give verbal instructions to a package deliver, get notified when someone approaches your door, and have video of all of these. As you can imagine it embodies all of the attributes detailed above regarding quality, hardware, software, ease of setting up (even with super goofy legacy DC doorbells), etc.

The Floodlight Cam mates all of these capabilities with a standard A/C powered direct wire light (the kind typically over a garage or on the side of the building). Same experience with software and integrates seamlessly. Just a no-brainer built in five years of execution. It will be available this Spring.

Voice and Cast

Everything without a screen can now be controlled by voice (I mean, Alexa).

Everything with a screen can now be a cast receiver (I mean, Chromecast).

By far the broad penetration, demonstration, and use of Alexa was the “big” industry news. While nowhere near as big, the rise of every screen as a casting receiver is a second example of a new “runtime” making its way broadly into the full range of products.

For new products at CES there are always two things people want to accomplish at CES — show the product fits with a broader context and connect to a strong, existing ecosystem. In past years we have seen things like Built for Nest, or integration with Facebook as ways to demonstrate this. So before we get too excited and declare the future of computing to be voice or Alexa as the next platform it is worth thinking about the motivations for such rapid adoption.

On the one hand, a company that makes a new device to be “controlled” loves the ability to easily and cheaply add voice. As we know from the Alexa Skills SDK this is not a massive amount of work and with Echo the best selling holiday electronics on Amazon this seems like a great bet. It certainly makes for a fun demo (although it doesn’t work too well on the show floor!)

Because IoT devices don’t have screens and you don’t want to access the app all the time (especially for demos) voice makes a compelling way to show off features. I’m a little skeptical of all the refrigerators being when you order food, but certainly hands-free control of cooktops and ovens seems perfectly reasonable.

I am long-term more than bullish on voice control and of course an Echo user. In the short term, I think collectively things are moving a bit ahead of reality. Learning a voice vocabulary that can work in practice is actually pretty tricky. The technology still does not handle complex asks beyond canonical 1:1 mapping of commands, with the twist that modern voice reco can handle more voices and more linguistic ambiguity. I believe we’re very early on this and a lot can happen between now and the voice future.

As an example of the penetration of Alexa, here is a giant display from Ford Motor that outlines using Alexa from within your auto. They built this into Sync3. It is great to see this but I have to admit the utility of this sort of stretches the imagination a bit (and so does Sync3).

As an aside, there were dozens of voice controlled solutions as part of music speakers or general IoT devices (desk lamps were a favorite) that did not rely on Alexa. These tend to use the ML-based solutions from Microsoft or Google to roll their own voice processing. This is an example of where it seems like there is much value to be added to encourage more use of these platforms or more for Alexa to do to be preferred over low-level roll-your-own.

Price/Value brand Cowin with a WiFi speaker with basic voice response capability.

On the cast side, specifically Google Chromecast technology, calling this out is just part of my own desire to keep code off of devices that should not have code. Rather than have an app platform for every streaming service on a screen with apps written differently than on the massive duopoly, I’d like to be as codeless as possible. I’d love to have nothing but power and WiFi for screens that aren’t full implementation of the secure, robust, manageable, updatable, mobile OS. I especially don’t want apps on a device that is not being actively managed for those qualities by the device maker, and we know TVs last a long time and exist in a highly fragmented global market that is exactly the kind of thing that is difficult to manage.

In addition, the use of casting solves the hard problems of complex interfaces on far away screens, dealing with remotes to hidden tuners and DVRs and so on. While there is ample competition from AppleTV I just don’t think the future is another box connected to a display, especially as displays become even thinner and easier to wall mount and WiFi continues to become even more reliable and faster. I also don’t think. All content/channel, box, or screen makers will be great at making and updating apps for a box.

Both Google and Amazon did a great job gathering support while not even having their own presence at the show!

Auto

I am very excited by what is going on in Autos. I’m intrigued by how CES has transformed the weird (and super tacky) North Hall from “car stereos” to have massive booths by the world’s largest manufacturers.

At the same time, this is very much a “vision” and “intention” part of the show. Even for an old-timer it is very hard to discern what is being done and when and by whom from essentially posturing. All the big progress here were really announcements of partnerships, proclamations, and commitments.

This year completes the re-branding or re-positioning started last year by these companies as they are “all in” as “mobility” or “movement” companies and not “car companies”. The American ones still make trucks and really care about them.

The biggest winner at CES is easily NVIDIA. Qualcomm also has a big presence. After that there are dozens of companies vying to be part of the supply chain with key technologies. It is likely there will be a wave of consolidation (either into the auto companies or into existing large scale supply chain partners).

IoT

By IoT I mean all of the gadget, appliances, devices that do things we already know how to do but now do them in a connected way with an app.

For many people this is where a lot of jokes come from at the show, and that is to be expected. There’s a tendency to view these products as luxuries or unnecessary.

Even if you are a big fan, these can be frustrating. Often enthusiasts quickly latch on to the potential but that races ahead of the reality. With IoT what enthusiasts have been trying for years (going way back to the idea of “home automation”) is not just a fancy way to do one thing but stitching together many devices into scenarios or whole house connected “macros”.

Together these perspectives create a climate of disappointment for IoT. For me this is a bit like being frustrated at early digital cameras, early PCs, even early electrical appliances. Most everything with a motor came from a manual tool and then turned into something with a transistor over my lifetime. Throughout this transition, it always seemed to be the case that the older ones “worked fine” and the newer ones were more expensive, less reliable, and overly complicated.

Then they become normal. Things that were once luxuries become necessities, even if you don’t really “need” them. Also new scenarios emerge — motorized window shades seem silly until you can now have shades for windows far out of reach.

At CES this year, there was little new in a broad array of sensors and automation to report. What was visible though was a maturing and improved “productization”.

Home security has always been a big area. In past years of CES we saw components such as web cameras or phone controlled locks. Even just last year we were only seeing cameras and locks as separate companies but connected to the internet. Now we are seeing the natural phase of consolidation and maturing.

Across the floor there were many complete systems for home security — no longer was this done as partnerships that really don’t work because designing shared cross-industry APIs for these scenarios that are always changing didn’t work back when it was X10 or later incarnations of this.

Blink Cameras were announced last year and were pretty amazing. It is a security camera/motion detector that runs on batteries so you can stick them anywhere (there’s a base station you connect to your home network). This is another execution example as now Blink is a broader home security system including a 4G base station (no internet needed) as well as a suite of perimeter tools. These are not all the time full video feeds, but motion based (much like the Ring doorbell).

When you consider the miniaturization, connectivity, and battery life that come together with sensors and cloud these systems are pretty mind blowing. With no special skills in an hour and with just a step ladder and a phone it is now possible to establish a home perimeter and monitoring system. No appointments or anything.

An incredible story of execution and of avoiding disruption can be found in home locks. When the first connected lock appeared it was still very early and challenging to get up and running. Several generations later these are easy to install, part of the home security ecosystem, feature-rich, and surprising to some also offered by “legacy” lock providers. Kwikset has a whole line of locks now including full touch screen versions, all of which work over WiFi and support multiple codes, bluetooth, and more.

Speaking of legacy companies, I loved this phone connected Safe from the “legacy” alarm and safe maker FirstAlert. It is the same home fire, water, break-in safe that you can buy on Amazon and have delivered by Prime (seriously that is cool for a heavy safe). This model is unlocked by voice or by a code on the app with multiple codes per household. It also tracks when it was opened and even has a motion sensor in the safe to notify you if the safe is moved which I think is super cool. It has a USB port for backup battery to unlock by code even if your phone battery is dead and of course a cylindrical key for backup. I wish it implemented two-factor authentication.

For monitoring yourself and securing yourself, there were several second generation products that now leverage your mobile phone to provide personal security (and an assortment of personal health telemetry depending on the product). I really like that Nimb ring (below, in charging station) which makes it easy to send “911" notifications via your connected (Bluetooth) phone and a cloud service by simply applying pressure to the ring. Motiv and Magnus are two other rings that were showing amazing combinations of features in a very tiny form factor.

On the home appliance front perhaps the most amazing transformation happening is the dominance of the space moving to the two main Korean manufacturers, Samsung and LG, steadily over the past few years. It is incredible to see this change. There are a lot of chuckles to be had with connected home appliances as they search for scenarios. The jokes write themselves. But some are super cool — certainly telemetry and diagnostics, even if they can just be sent to a phone when scheduling a repair is definitely a time saver (if you’ve ever had a repair person show up for an appliance you know they rarely have the part in the truck).

There’s a battle royale going on over two-in-one washers and dryers. LG came out with the double-washer plus large dryer two years ago. This year Samsung one-upped LG with dual-double — a washer with two washers and a dryer with two dryers. They are connected too, so you can get notified when the load is done, which is kind of useful (though “Alexa order detergent” might be less so).

Solving a problem everyone has in spades these days, Ucella offers a product that makes it easy to accept packages at your house when no one is around. The collapsible device attaches at a place where delivery agents can see it. When they make a delivery they scan the barcode as they normally would which then unlocks the device and allows it to be placed in the secure (expandable) box. The service does a lot of cloud-based connections to the delivery services to enable full tracking when using this destination. Ucella is Silicon Valley startup. Sign me up.

While IoT in the consumer space has been criticized for being more novel than useful, the industrial and business space is going strong and reinventing many categories (like in the GE commercials!). The show does not have a lot of industrial IoT because those are highly specialized but I did see a good example of the kind of product that is already in widespread use. This SPC-Smart Power Carrier is a motorized “dolly” but what is cool about is that it uses a smartcard or phone to authenticate the human user (or otherwise lock the device and load in place) while also using WiFi to track the load and weight. When you can’t use a robot to move heavy items around a floor or warehouse you can still have the benefits of automation for efficiency, safety, and protection.

Mobile OS/Chip/Devices

In one sense, every device at the show is a mobile device because everything is built on top of the ARM-based mobile ecosystem (for the devices, not the cloud services/data center). The one exception are the tethered AR/VR glasses which nearly every single one will tell you has a goal or is in process of developing untethered devices built on ARM.

There were some interesting devices on display. The Samsung Chromebooks that were announced—confusingly the Pro and the Plus. Even more confusingly, the devices are essentially the same except one has an Intel chip and the other ARM, but both run ChromeOS with the capability to run Android apps which is new (currently in devchannel testing in a few devices, I’ve been using one). The devices are packed with touch screen, long battery life, and solid screens. There’s an accelerometer to bring parity with phones, but you have to admit swinging around a laptop is weird.

The advantage that ChromeOS has in running Android is that the whole Android OS/runtime is on the device and apps just run totally natively—there’s no emulation or virtualization. It is clear that Android apps will need to be updated to function properly in the presence of a keyboard and trackpad—this is not seamless and my experience lately has been flakey. I totally get why you’d want Android apps (more stuff is better), and understand why it works (the whole of Android is there and the hardware is phone-like), but am not sure if it ultimately undermines the value proposition. Android is not free of challenges in the environment where ChromeOS is thriving today. What I am unclear on is how this capability will maintain the rock solid, secure, reliable, easy, appliance experience in today’s ChromeOS. We all know that many services are becoming app-first or even app-only, and at the same time the Play store is not free of socially engineered malware challenges nor is Android free of device management/security challenges.

The Huawei Mate 9 is a new Android phone. What distinguishes it are two things. First it will be the first phone to come preinstalled with Amazon Alexa, which is really just adding the app and providing services/skills setup by default. Second, it comes with a very fancy double lens Leica designed camera lens (not a Leica sensor). The camera app has been updated with a feature-rich implementation of Apple’s Portrait/Depth mode. You can adjust the aperture after the photo is taken and you can also adjust the point of focus. The de-focus effect is a stylized blur as Apple has done, not a true bokeh.

Long range radios have become a mainstay for outdoor activities, safety work, and large scale group activities. For some time, the high end ones have had fancy mapping, text messaging over their own networks, and other phone-like features. The ShowPro takes this a step further and also adds a camera for photo/video and sharing on the radio network. Essentially you have all the capabilities of a “phone” but running over the radio network and PTT. Since in many circumstances your phone won’t be usable, this saves on carrying one extra device.

TV

In keeping with the theme of execution, this year saw OLED finally emerge from the labs and demos. Earlier last year LG started selling OLED TVs up to 65" and this year sets from Sony, Samsung, and more will hit the shelves and get larger.

A key part of this execution is that these flagship TVs are now all 4K, HDR, and include DolbyVision for color (yes, the sound company is also doing color now). All of this was discussed and shown last year but is mainstream this year. For pricing expect these to be a 3–5X premium over LED depending on size, but LEDs are so (relatively) inexpensive now you still won’t pay anywhere near what was paid when flat panels first came out. It is always amazing when Moore’s Law is applied to screens too!

Samsung showed an extremely nice proprietary wall-mounting system that keeps the TV entirely flat (better than most pictures I hang!). It also has a fiber optic cable to a separate jack-box which allows for a clear ultra-thin wire to hang down the wall (assuming you have mains behind the screen this is very subtle). I’m not a big fan of these proprietary cables in general but if you have an equipment stack this is nice.

Sony showed his 4K/HDR projection TV, the VPL-VZ1000ES, which can project a 100" image from 6" away (demonstrated last year). It is heavy and costly though and for less money you can get to nearly the same size with a very nice OLED screen and <1" of wall depth so I am not sure of this. It is a marvel though!

I am fascinated by the developments in cord-cutting PLUS casting and how this could really change what a TV is. All the makers are still pushing “smart” as a key part of TV. This is a challenging value proposition for them as they are not set up to keep these proprietary apps up to date. Their product lines are complex and fragmented and they generally don’t like to support models for many years. I’m pretty interested in thinking about how a TV could just be a cast service with no tuners and no apps (or maybe the cast receiver code, which is basically hosted by Google). Since finding and “tuning” to shows is so much easier with a phone or tablet, it is seems that we are on a path to free of both boxes and dongles as well as an app-platform on the TV.

There are both curved and straight TVs. 3D is nowhere in sight.

My favorite TV though was this crazy huge Sony “canvas”. It is made up of little squares that seamlessly stitch together with image processing. It is an insane image and super cool for industrial uses.

Drones

There were a lot of drones on the floor. The supply chain has done a great job making it easy to get things into the air. Most all of the drones on the floor are copters of some kind, with minimal range, minimal payload. These are basically drones for consumers or as “toys” and I don’t see that changing, except for regulation (in the US at least) becoming much more strict.

Drones show that software is the gating factor for this category. Flight management, sensors and imaging, image processing, and a whole host of other safety and reliability features all require more software than these companies can make. That is why drones are so clearly split between consumer and industrial. There’s very little by way of industrial drones at CES.

DJI, with a big floor presence, is doing a very solid job of positioning in the middle as “prosumer” building up from consumer and on a path to be professional.

The big drone news during the show was their acquisition of Hasselblad, the legendary optics/camera company, clearing indicating a move up the industrial stack. We’ve seen this playbook before as it is how Windows became a server or how Canon came to dominate professional photography.

If I had to pick a drone to highlight, it would be the $600 portable, foldable, follow you around hovering camera, The Hover Camera Passport from Zero Zero Robotics. It basically follows your smart phone from 100 feet up. I like it because it is simple and the blades aren’t exposed (which has implications for flight in the real world, just not in this toy world). It works from a simple app and except for escaping from the cage once on the show floor seems rather fun.

VR/AR

After a big start with major platform players leading the charge with VR/AR, this year saw an explosion in activity in the space. My view is that we’re seeing a lot of activity that is super early, but required for innovation. We are in a post-invention, but pre-platform/pre-consolidation phase.

On the floor were too many headsets, motion tracking, eye tracking, audio, content authoring, and other technologies and tools to dive into. This reminds me of the “multi-media” era of PCs before the world settled on some formats, players, and tools.

As mentioned, most every device (numerically not by share) is tethered to a PC or not particularly mobile yet. At the same time, everyone wants to be free of a PC.

The raging debate over VR “versus” AR will continue. For me, the far more interesting question will be who/how these technologies break out to go beyond gaming/entertainment. I’m fascinated by the fact that we are seeing innovation in this space while at the same time debating whether we can use phones in theaters—the second screen is so critical to enjoyment/engagement for entertainment these days! Yet, the immersive experience of VR is so mind-blowing it might be a paradigm shift that returns us to immersive entertainment.

AR is always shown in education and industrial settings. One popular demo was using AR for medical education instead of learning anatomy on real cadavers or for use in repairing industrial equipment when the exploded diagrams can be visible while looking at the system. These are definitely interesting examples but require incredible sophisticated tooling.

There wasn’t any visible low-end mobile app-AR like we saw with Pokemon Go. I still expect to see more of this in apps like home remodel, repair, and so on.

AR/VR represent the best of CES “ingredients” for those working in the space.

PC

There was definitely a lot of experimentation going on with PC form factors. Some are talking about this as a return to excitement of the PC. If you’re using a PC and looking for a new one you will no doubt have more choices, though the open question is whether you will be doing anything differently once you have a new form factor.

PC Gaming which is enormously popular is driving high end compute+graphics form factors. There was a lot of excitement around the Razer Project Valerie—a laptop with three fold-out 17" monitors and a high end CPU and discrete GPU. It is some pretty impressive engineering. Razer has a good track record in shipping their “Project” PCs when they say.

The LG Gram continues to be the thinnest and lightest around. It is fun to hold on to and think about carrying around. It uses a proprietary power connector though which means you can’t leverage a phone charger for weight savings. There are a number of PCs that come close to this weight.

Dell, HP, and Lenovo all announced new PCs including All-In-Ones, convertible, and very thin/light laptops. It is great to see the prevalence of touch screens though the renewed focus on desktop keyboard and mouse from Windows makes this somewhat limited in utility.

The HP Envy Curved A1034 is an all-in-one form factor with a 34" curved “wraparound” monitor running 3,440 x 1,440 pixels and full Core i7 CPU and a stylus. Definitely an attention getting for that command center look!

The Huawei Matebook is a pure tablet running Windows 10 with a 12-inch 2160 x 1440 running Core M (effectively a last year MacBook). It has a minimalist keyboard pleather portfolio. The hardware is well-executed at 278.8mm*194.1mm*6.9mm weighing under 700g. There’s a charger/hub with RJ45 and USB 2.0 ports integrated with cord storage.

There’s always a small vendor making some super tiny and cool looking PC. Here’s this year’s 3x6x0.5" PC with everything in the box, just add a display. It is the Dsuome DS152F Fanless Mini Desktop PC by YMTronics.

Look Surface clones! In all seriousness, OEM execution has effectively caught up and raised the bar, though their distribution approach and model line complexity means they are not always competing effectively with Surface.

Health

Health IoT for consumers did not make as much progress this year. My sense is the majority of progress is focused on hospital/professional use where there is significant progress on connected/integrated devices.

The core challenge for consumers is to provide reliable and actionable health information. The exhibition is still dominated by too much non-scientific or highly variant measures of important telemetry. One concern is that these devices actually increase medical costs and strain the system more as people worry more than they should about things that are “normal”. There are also too many devices that are, politely, lacking in scientific rigor (hint, blinking lights do not make hair grow).

The phone as universal host for sensors and connection to cloud data is a revolution. Here are couple of devices I saw.

The EyeQue vision tester is one of many devices to take the pain out of the “better/worse” dynamic when visiting an eye professional (or in markets where access to existing industrial equipment is too difficult). EyeQue attaches a scope to an eyephone and does self-adminstered reflection measurements of vision, with or without corrective lenses. There are several variants of this scenario on the market and it is an important one for emerging markets for sure.

As with home security, this year saw execution in body vital sign measurement bringing together several devices that were separate last year. Two examples demonstrate this execution. MedicalCenter by Croise measures body temp, fat%, ECG, HRM, BP, and has a glucose strip reader all in one integrated device. The execution challenge here remains BP as this is an area where there have been decades of work trying to measure without a cuff and so far the other measurements are too unreliable.

TytoCare takes a very different approach. It is designed as a suite of tools that are integral to a telemedicine service. I think this is very interesting because it offers the patient a way of giving a physician more information than blurry phone pictures or text descriptions of routine challenges. One device with various attachments measures 1) heart auscultations, 2) lung sounds, 3) stomach/GI sounds, 4) throat, 5) ears, 6) skin, 7) temperature, and 8) heart rate. There’s a pro version and home version.

The WHILL is an intelligent electric vehicle. The WHILL is a modern chair using all the latest in battery technology, wheels, and more. In addition it is connected to a phone which provides all sorts of telemetry for understanding how the vehicle is performing. I watched several people test it out on a rough terrain track they set up and the feedback was very positive. Cost, service/support, availability are key assets they will bring to market as well.

One connected device related to healthcare that I saw a lot of were natural drink makers — cider, orange juice, frozen, coffee, and of course beer among many others.

Imaging

Imaging is both an exciting area and very much an area that has been disrupted by cameras on phones. In the 1990’s the consumer digital camera debuted at CES (VGA resolution, $500, with no printer, no internet, and very expensive storage). In 2000 the film to digital cross-over happened for consumers.

Mostly the imaging at CES is now for professionals and so this year was a focus on 4K/HDR video.

Canon showed off two previously demonstrated cameras. One is a zero-lux image capture (i.e. capture images in darkness). The other was a spectacular 250 mega pixel camera that had an insane video of a person on the top of the Eiffel tower waving captured from 12 miles away. The sensor is housed in a box that uses fiber connected to storage where it takes about 10 minutes per frame to post-process.

Both if these were technology demonstrations. They said “looking to see what customers were interested” as if this was not already flying around in satellites for the military!

There were a half-dozen grips for phones to add gyroscopic image stabilization to camera-captured video. Super neat as the sensors become even better. Image stabilization is now required in DSLRs and likely will make its way integral to a phone, as software.

Alternative spectrum cameras have always been highly specialized. Flir brought thermal and continues to innovate (with a lot of booth demo discussion over thermal versus other spectrum). eMagin showed off an attachment for night vision capture on an iPhone that uses the phone’s camera to aim and then captures night vision enhanced image with its own sensor.

Kodak continues to rebuild itself. At the show they demonstrated an Android phone with their lens and camera app (not unlike what Leica did with Huawei). They also showed a 4x6 transfer printer which was very well done with an iPhone (or Android) dock for connectivity to phone. These can be nice to have for some sense of permanence rather than going to CVS. By the way, Kodak also reintroduced Ektachrome 100 film, the film favored by National Geographic for decades.

More…

While many of us are already using WiFi Mesh solutions, the competition in the space took a substantial and predictable tick up at CES with introductions of mesh solutions from all the consumer networking providers (Linksys Velop, TP Link Deco, Netgear Orbi). Looking forward to continued roundups and speed tests. The visible similarity of all of these to each other is noticeable, except some are pucks and some are obelisks. My favorite observation—Deco used USB C connectors as the power adapter which is laudable.

The speed at which USB-C has taken over is impressive and helpful. If you’re into dongles and connectors you can find all sorts all over the show floor. There were dozens of hubs and docks, especially for MacBooks. There’s a lot of complexity people are starting to realize when it comes to the overloaded connected relative to power, traditional serial protocol, and video/thunderbolt. Also noteworthy, most new monitors are adding thunderbolt USB-C.

After about a decade of trying, I think wireless HDMI works. IOGear had both a long range 300ft pro solution and a home solution with a 5x2 matrix. If you have been trying to get HDMI to travel over CAT5 this might be a better solution. You can put a CATV/SAT tuner in one place and then tuck a small box behind a panel with an IR receiver and send it to 5 different panels.

Bonus: CES History

As this was the 50th anniversary of CES there was a mini exhibit to show off some of the big announcements over the history of the show. Below you can see some of the firsts I thought were pretty cool: Moto StarTac, Sony Walkman, Timex Datalink, Sharp Wizard, KwikSet lock, Compaq 386/LTE, Magellan GPS, TiVo.