AWE2016 Santa Clara June 1-2.

Thoroughly enjoyed AWE 2016 in Santa Clara.

Presence from AREA.org was one main thrust. AREA.org presented use cases for AR in utilities and manufacturing, with emphasis on asset augmentation. Another main topic was introduction of latest AR/VR gear: notably Epson Moverio BT-300 - high resolution - but taking a different path from sensors on Hololens. Would be remiss without mentioning Vuforia from PTC. Found out about AWE from Vuforia media campaign. Learned more about AWE on website and from pervious show videos. Got my AWE pass via Twitter.

Best AWE giveaways: Reality tops. VR cardboard headsets.

Best AWE takeway: AR for use in asset augmentation... reporting location, status and what needs to be done to foward the business applications.

As personal project: Always strive to "do whole trip without automobile". It went thus.

MBTA to work and then MBTA Blue Line in afternoon to BOS.
Jet Blue BOS to SJC.
VTA#10 Airport Flyer.
Metro Airport on Green VTA - Mountain View (actually took Blue - Alum Rock and changed at Tasman) to Great America (in front of San Jose convention center where AWE was being held). Then walked about a mile south past Great Anerica to Hotel Avatar.

Was very pleasant weather the whole time in San Jose. Walked out and back each day to convention center. Hardest part was that all food options closed at evening arrival at Avatar. Got a burger at nearby In and Out Burger. Got groceries at Walmart for the balance of days. AWE had no luggage holding area, so had to go back to the Avatar afternoon of departure to pickup bags.

Walked to Green VTA  - Winchester - to Metro Airport.
VTA#10 Airport Flyer.
Jet Blue SJC to BOS redeye. Packed flight. But able to sleep.
MBTA home, Blue and Orange Line, for a nap, referesh and then headed to meeting downtown with International visitors to Cimetrics.

Signatures. Detection. Pattern Analysis. All Kin.

Much of what we do each day as humans is pattern matching and signature detection. Cimetrics has this at its core too. It appears everywhere....

Signature detection is central to image recognition. Witness Tensor Flow, Ditto Labs, and others. Signatures arise in audio and RF (and even home brew RF using RTL SDR and others). And in power and water line fluctuations. Witness Norford and Schweppe at MIT. And their follow-ons - Zenzi - and even more currently Sense (at Bolt) - as opposed to Sense.io for big data analytics - not unrelated. And both too close in name to "Zensi" and "SensEye" and others.

See recent Cimetrics blog post on signatures.

Replace, Repair, Upgrade, Other Pathways.

You have had "old reliable" for ten years and you are unsure about whether to replace, repair or upgrade. Usually issue comes down to cost benefit analysis. And are there other options which you have not even considered?

Even if one has decided to do something like Continuous Monitoring or Lifecycle Management of the unit (because one has a programme for all units), the choice is still hard.

One craves data to inform decision, but a study or upgrade may be needed to even get data. That costs too.

A. Get a study. Or hook directly to remote monitoring. Externals are involved and that can cost. For too many reasons to list herein... that often will not work.

B. Replace it with modern gear and start fresh. Hard to argue with that. And if it can be gotten through - golden. But one will never know what was thrown away... and it will be questioned.

C. Repair and Upgrade. This is a trick option. When one is in the process of doing repairs, it always makes sense to do an upgrade. But you are at the mercy of the contractors who do this work. However often it is the best value. 

D. Give it a born again experience - warts and all. This last one was once as expensive as all of the above, and at least as much trouble.

What if you could magically turn dumb old reliable into a smart new system with self diagnosis. Not saying that we fix old reliable yet, or even have it mend its ways. Just make it talk. Just make it self aware.

That was just not an option as little as five years ago. It was expensive. Only the manufacturer or installer of old reliable could do such magic and their incentive was always to replace and repair first and ask questions later. Upgrade to self awareness often gets value engineered out of the process.

But think about what

• - NEST does for your furnace and AC system.
• - RESA Power Systems does for old switchgear.
• - Any OBD-II monitor can do for your car (insurance).
• - Eaton PXG600E does for a Westinghouse metering system.
• - Roku, Chromecast (or similar) does for your home entertainment system.

The age arises where you buy a "rebirth" box for old reliable: Hook the box onto old reliable (with sensors and wires to what little smarts old reliable has). Hook the box onto your existing network (and we all have networks - even if we do not think about them). Analyze old reliable with Internet tools.

Past ages made such a rebirth box expensive because:

• A. The box hardware was expensive. A PLC is thousands.
• B. The setup of the box was expensive. System engineering is just plain expensive... again in the thousands.
• C. Maintaining the box was expensive. The ecosystem of data exchange and support required massive organizations.
• D. The box (because of hardware, setup and ecosystem costs) was made to do everything possible - but not one thing well.

The key is in the economics of chips, simple interfaces, auto-configuration, and supporting ecosystem that have come from  the explosion in mobile devices. And the mantra of doing what it should do well is finally coming.

I like the idea of reuse. It turns the installed base into an opportunity instead of a liability. And it is good for everyone. I like the idea of a smart device. David Rose - Enhchanted Objects. I like the idea of this being magic.

Programmable Logic Controllers, PLCs, Reliability.

PLCs are not about all the long list that gets hauled out about their speed and ease of programming and all the issues of "how they work". PLCs are about reliability. PLCs are expensive... for reasons that many are too late in understanding.

You can make your own PLC if you work out how to get the hardware and ecosystem up to or better than the standards for existing PLC reliability. Most will never understand what this requires until they fail. Nor will they have the self awareness to get past their own flaws and look outwards (and maybe even bypass failure?).

Am I challenging you? Yes. Am I challenging you to "self sufficiency" or "rugged individualism"? Look at how ecosystem plays to reliability and then ecosystems like the open source and maker movement and then decide.

Are PLC use cases what is driving the massive reduction of costs in underpinnings of their function?  Spoiler: It is the glass screen in your pocket, and the search bar everywhere.

"Come now, let us reason together". You want to build one? Yes you do! *grin* Let's see how.

Docker IoT Meetup VMTurbo NVBOTS Resin.io.

Docker Meetup at VMTurbo Boston offices. Great presentation by NVBOTS. NVBOTS has a turnkey lightly managed 3D printing system aimed at organizations without expert staff, like schools and such. NVBOTS on the road to make 3D printing go down the self serve pathway that laser printers have long since travelled.

NVBOTS are using Resin.io. Resin.io has heterogeneous agancy management system for things like Raspberry Pi, Intel Edison, Beaglebone, etc.  Then Docker for DevOps. All together to manage your IoT fleet.

IoT So Far.

Have been following IoT now for three years. And it is still too soon to say how it will all play out. Here are a few observations though...

The billions of device predictions are correct in their own way, but one can credit the original RFID IoT banner carriers with making that prediction twenty years ago. Logistics (having perhaps the smallest motes of IoT and widest field) has always been a great place for automation, and it is no where near over. And if one allows handheld devices (phones and tablets) to fall under IoT (some do not) then clearly.

Sensor and analytics players contend (though admittedly it is self serving *grin*) that the only really new parts of IoT are sensors (new ways of sensing) and analytics (new ways of understanding). These can both be covered by the two different definitions of "seeing" and thus the single banner of "Seeing the Unseen". And they are the extreme ends of the chain that one can call IoT.

What of everything in between? Contentious, but not innovative. Technology in LAN, WAN, databases, aggregation, platforms, visualizations - are all well understood.

So some idea of who will prosper? And for what..

• IBM IoT. Because they are IBM.
• AWS IoT. Price. Open to all comers.
• Microsoft Azure IoT. Compatibility.
• Google. Search.
• Apple. Style.
• Xively. Security.
• Thingworx. Mashup. Design.

And some in well entrenched verticals who roll their own:

• GE Industrial.
• Auto manufacturers.
• Aerospace.

Or who are new and forward looking and for whom it will "just be"... in AI, AR, VR, Robotics.

Fast Food Combo Deals.

Noticed that fast food places had many combo deals running. Heard it this morning on local NPR as well.

Here is a piece from a WWNO.

Have been using combos to build a larger (healthy?) combo. Put chicken nuggets or the like, with a portion of a mixed green salad bag, and some dressing (store brand from Market Basket, Wegmans, Shaws or Stop and Shop will do nicely). If fast food could handle the lettuce logistics, then such a combo could easily be their combo. Wendy's small salad comes close. Maybe a Leafy Green Machine from Freight Farms out back of every location could provide the goods?