Big Data Pays a Visit

Today a coworker of mine came up to my desk and asked, “Do you want to see something weird?” “Sure!” I enthusiastically replied. Who doesn’t want to see something weird? So I went to her desk and there she showed me her FaceBook page, where there was an advertisement for a desktop wire file sorter – the model I had just bought for the office. She had never seen one like that before I’d put it on top of my cubicle wall a couple of days ago, but now here it was appearing in an ad on her FB feed.

I had done a search for it on (didn’t buy it there, though), but she and I are not FB friends, and so the connection between us isn’t obvious. Then she scrolled right on the ad, and up came a nightstand – the very one that had been sitting in my Amazon cart for a few weeks. I’d dropped it out of the cart a couple of weeks ago, but here’s the interesting thing: I searched for the file sorter using my business account, and the nightstand was in the cart of my personal account.

“Clearly there’s a connection between you and me,” I said.  I told her about Big Data, and how big companies like Amazon and FaceBook are collecting all kinds of information about us, who we know, what we like, and so forth.  I also told her how, a few days ago, I created an Instagram account so that I could see the wedding photos where another coworker was a bridesmaid, and again, that other coworker and I are not FB friends and have no other connection but work.  When I created that account Instagram suggested some people I might want to follow and amongst some random famous people was yet another coworker, one of our warehouse guys, with no other connection between us.

“So, the lesson here,” I said, “is to be careful where you buy your sex toys.”

EmDrive Pushing Forward

Here’s more news as a followup to “EmDrive – Propulsion Revolution” – NASA continues to test the EmDrive reactionless thruster with promising results, and they’re making progress with theory development to explain the otherwise inexplicable test results.  There’s a report that experiments show EM drive works in a vacuum, thus dispelling the leading debunking theory that the measured thrust was an artifact arising from internal thermal convection.  What’s more, the working theory helps explain disparate results between U.S. & UK & China teams.

This is exciting!  Read the latest here, although sadly there’s no word yet on what it sounds like.

EmDrive – Propulsion Revolution

NASA confirms: EmDrive works – propulsion without propellant. Yes, I’d say that’s OMG WTF big news!  According to the homepage FAQ, 1st generation will be suitable for spacecraft thrusters, and the 2nd generation (which assumes use of superconducting cavities) should be suitable for terrestrial ground vehicles.  I wonder if the EmDrive goes “weeble-weeble-weeble“…

X-37: The Spy Shuttle

Something curious happened earlier this month: the National Reconnaissance Office donated two perfectly-good satellite frames to NASA.  They’re still sitting on the ground and they have no instruments, but apparently our spy satellite agency has no forseeable need for them.  Not that the NRO cares, but NASA has neither the need for them nor the budget to do anything with them.  They’re probably going to sit around in a NASA warehouse for a good long time.  Why doesn’t the NRO just keep them around and then fit them with instruments and launch them as budgets permit or circumstances necessitate?

Now, I’m no spook and I have no insider knowledge but I believe the answer lies in the U.S. Air Force’s X-37B Orbital Test Vehicle.  This is the unmanned mini-shuttle that’s currently in orbit on a year-long test mission, and is scheduled to return this month.  Insofar as we can tell – not that anyone’s talking – the program seems to have been a success.

Current spy satellite design has some inherent drawbacks.  Once launched, they will hopefully serve for a decade or more, which is a bit of a mixed blessing.  The longer they continue to operate the older and more outdated their instrument set becomes.  Instrument performance tends to degrade, and the technology remains a relic of the age in which it was designed.  A satellite which has served for a dozen years probably has instruments that are fifteen years old, based on technology which is perhaps twenty years out of date.

Servicing missions were never practical; now that the shuttle fleet has been retired a servicing mission is entirely out of the question.  What the NRO is doing amounts to a paradigm shift: they’re going to build a mini-fleet of mini-shuttles, which can be fitted with a current-generation sensor payload and launched virtually on-demand.  They’ll probably keep a handful up at all times, scheduled to come down after a year or so to be repaired, refueled, and upgraded.

There’s been a lot of people doing head-scratching and fearmongering over the X-37 program.  Granted, it might have some uses as a weapons platform, but I believe the real advantages are in surveillance and espionage.  IMHO, the X-37 is the test vehicle for a new generation of spy satellites, which will enable us to cycle out old technology for new on a comparatively rapid basis, while extending the lifetime of the platform itself by enabling us to repair and refuel it on each mission cycle.

Update: Some interesting, related thoughts from “George Smiley” over at “In From the Cold”.

Modern Electronics Designed to Fail. Yay.

[by Mr.Hengist]

Interesting: “It is by no means uncommon” for modern IC integration design engineers to assume the fast obsolescence of the end product, and therefore IC environmental protections are not incorporated. Such devices will therefore have short lifetimes by design, and while that’s not a problem in market due to fast obsolescence, it means that you’re SOL if you want to keep such a device for an extended period of time.

It reminds me of the story of Henry Ford, who perused junk yards for his cars. He asked the proprietors which parts lasted longest, and then instructed his engineers to make those parts more cheaply because, by his reasoning, they were overbuilt.

Anyways, here’s the relevant quote:

The traditional functions of a semiconductor device package are to protect the die from degradation by the atmosphere and fan-out the electrical interconnects to the next level. Because of the benign environment in which most modern semiconductors are used coupled with short expected life through product obsolescence, the need for the package to provide environmental protection has virtually disappeared. It is by no means uncommon to see essentially package-less chips attached to circuit boards, with just a polymer covering over the exposed bond pads.