Saturday, September 30, 2017

End of September

It's September 30th, or as we call it around here: August 61st.  The end of September with only one sign that it's not August: the sunset getting earlier every night while sunrise gets a little later every morning.  To be honest, it's not quite August.  Yeah, it has been in the 90s if it's not raining, but over 90 doesn't ordinarily last as long as the high temperatures in August.  Accuweather's long range forecast guesstimate doesn't really predict a overnight low under 70 until Halloween Weekend, but nobody should trust a forecast that far out.  They do show some highs in the low 80s in just two weeks, which is marginally more believable. 

And we're ending the month much as it began, under what the Hurricane Center calls "Invest 99".  (For those who aren't familiar, an "Invest" is an area of investigation, the first stage in the sequence of development of a hurricane: Invest - Tropical Depression - Tropical Storm - Hurricane).  That only means a few more storms and a bit more wind than usual.
Tonight's weather shows the center over NE Florida, with a predicted 20% chance of developing through the next five days.  That area over the Caribbean is given a 0% chance of developing over the next five days. 

 Just think; it's just under 8 weeks from last Thursday will be Thanksgiving. 


Friday, September 29, 2017

Best Thing I've Read on Artificial Intelligence Taking Over the World

By link at an out of the way place, CNCCookbook.com, I ran into this cool article, "Camels to Cars, Artificial Cockroaches, and Will AI Take Your Job?" This is really a well done piece and I recommend you Read The Whole Thing. Best thing I've read on this topic in quite a while, if not best ever. 

CNCCookbook publishes the "Speeds and Feeds" calculator I'm using, GWizard.  The owner is a guy named Bob Warfield.  Bob's an interesting guy; he has founded a handful of companies in the software world, and I think he says CNCCookbook is his seventh company. 
Before I launch into my reaction to these all to common predictions that AI is right around the corner and will take all of our jobs, let me establish my own credentials.  Hey, anyone can have an opinion, but like everyone else, I think my opinion is better!

I have worked in what many would call the field of Artificial Intelligence.  I made the largest return I’ve ever made selling one of my 6 Venture Capital Startups to another company.  The technology we built was able to automatically test software.
Bob points out that AI has been riding the Gartner Hype Cycle for a long time. In last summer's Gartner summary, they put it near the peak.  For the third time.
In fact, all of these technologies are near the peak of being over-hyped:
  • Deep Neural Network ASICs
  • Level 3 Vehicle Autonomy
  • Smart Robots
  • Virtual Assistants
  • Deep Learning
  • Machine Learning
  • NLP
  • Autonomous Vehicles
  • Intelligent Apps
  • Cognitive Computing
  • Computer Vision
  • Level 4 Vehicle Autonomy
  • Commercial UAVs (Drones)
What's different about this time?  Is anything different about this time?  This time we have demos!  We have Deep Blue beating the world's chess champion.  We have Deep Blue beating the worlds Go champion.  Well, yeah, we had demos the other times, too.  In the last peak of the hype we had
  • Medical diagnosis better than what human doctors could do. See Mycin for prescribing antibiotics, for example.  It was claimed to be better than human doctors at its job but never saw actual use.
  • All manner of vision and manipulation. Blocks? So what. Driving cars?  Yeah right.  Turn ‘em loose against a New York cabbie and we’ll see how they do.  The challenge for autonomous vehicles has always been the people, not the terrain.
No matter how many autonomous cars drive across the dessert (talk about the easiest possible terrain), they’re nowhere until they can deal with stupid carbon units, i.e. People, without killing them or creating liability through property damage.

By the way, despite awarding numerous prizes of one million dollars and up, so far the DARPA Grand Challenge has failed to meet the goal Congress set for it when it awarded funding–to get 1/3 of all military vehicles to be autonomous by 2015.  But the demos sure are sweet!
  • Computers have been solving mathematical theorems for ages.  In some cases they even generate better proofs than the humans.  Cool.  But if they’re so good, why haven’t they already pushed mathematics ahead by centuries?  Something is not quite right with a demo that can only solve theorems already solved and little else.
  • Oooh, yeah, computers are beating chess masters!  Sure, but not in any way that remotely resembles how people play chess.  They are simply able to consider more positions.  That and the fact that their style of play is just odd and offputing to humans is why they win.  What good is it? One source claims Deep Blue cost IBM $100 million.
When are those algorithms doing to genuinely add $1 billion to IBM’s bottom line?  Building still more specialized computers to beat humans at Jeopardy or Go is just creating more demos that solve no useful problems and do so in ways that humans don’t.  Show me the AI System that starts from nothing and can learn to beat any human at any game in less than a year and I will admit I have seen Deep Skynet.
One of the marketing gurus doing AI demos says "all we gotta do" is wait for computers that are about 100,000 times faster than what we have, and then overstates Moore's law to say we'll have them in 25 years.  If computers get twice as fast every two years (and the actual clock speeds plateaued around 2006 and aren't going up, but let's ignore that and say we get twice as fast due to architectural improvements) that takes 17 cycles or 34 years for computers to get 100,000 times faster.  It's gonna be a long time before we have Hal "open the pod bay doors".  Besides, I have evidence Moore's Law died in 2012 so we may never get there.

Think about the problem the other way, though, if a computer 100,000 times faster would be as good as human brain (and we still have serious gaps in our understanding of just how the brain works - including whether or not our brains do quantum computing), what would be the comparison to today's computer?  Could we get useful work out of what we have?
So we need artificial brains that are 100,000 times more powerful.  In essence, we can compare today’s AI to brains the size of what cockroaches have.  Yet, we’re worried they’re going to take all of our jobs.

Are you in a job that a cockroach could do?  I hope not.

So far, I am not aware of anyone having harnessed cockroaches to do their bidding, but they are cheap, plentiful, and just as smart as today’s AI’s.  Maybe smarter if their brains are quantum computers too.

Maybe it would be cheaper to spend billions learning how to make cockroaches useful?

I don’t know, but we don’t even seem to be able to make much smarter animals useful.  Are there dogs running machinery somewhere in China?  Is a particularly adept German Shepherd behind the latest quant trading engine on Wall Street?

Nope.
Decades ago, I read about a drug company that trained pigeons to be Quality Control inspectors on their production lines.  The gelatin capsules coming off the production line would sometimes stick together, so you'd get two tops or two bottoms stuck in each other.  The production inspectors would watch the molding machine's output on something like a conveyor belt and pick out the defective gelatin caps.  The humans would get bored with such a menial task, their attention would wander, and defective capsules would get through.  The pigeons found it interesting enough that they paid more attention.  As a result, the pigeons were actually better inspectors than the humans - they found 99% of the bad capsules.  The only reason they didn't make the pigeons permanent inspectors after this experiment?  They were afraid what the competition would say about them if they discovered they were using pigeons. 

Can you imagine cockroaches on the production line doing this job?  Maybe you pay them with the gelatin capsules they reject.  And can you imagine what the competition would say about having trained cockroaches inspect the medical capsules? 

Again, let me leave you with a quote I've used before, because I think it's great.
William Bossert, legendary Harvard professor, summed it up by saying, “If you’re afraid that you might be replaced by a computer, you probably can be—and probably should be.” While it may not be comforting, it could be a wakeup call for continued education.




Thursday, September 28, 2017

Since Tax Reform is in the News

I posted this back in June, but it really stuck with me and became one of those quotes I can't forget.  It's from David Stockman, Reagan's budget chief, who said,
The income tax has been slashed so many times since 1981 that it’s no longer a broad based societal tax; it’s a kind of luxury tax on upper income salary earners and the small share of households which garner most of the capital income from dividends, interest payments and capital gains…
Stockman went on to point out that 60% of tax filers accounted for 5% of tax revenue.  In addition, 35% of tax filers – more than 52 million filers at the bottom and middle of the income ladder – didn’t pay a single dime of taxes after deductions, exemptions and credits.  I think Stockman was sloppy with his wording in saying, "the income tax has been slashed so many times".  Tax rates go up and down regularly.  I think he means the number of people paying income tax has been slashed. 

Look at what he's saying.  The top 40% of incomes pay 95% of tax revenues, and the bottom 35% don't pay a dime once all of the deductions, and all other systemic ways to avoid paying tax.  The top 1% of earners pay 39.5% of all collected taxes.

From what I'm able to see, this isn't really a conservative plan it's more a populist plan.  From a populist president and party that aren't conservative, so what a surprise, right?   I hear the president insisted on leaving the highest tax bracket for the highest income earners where it is, at 39.6% so there's no "trickle down" aspect to this.  Supposedly Trump said that people in the highest bracket can afford that 39.6% rate, and he insisted it stay there.  Maybe that's his way of fighting the Evil party screaming that it's a "tax cut for the rich!!"  Maybe he honestly thinks a 39.6% rate is a good thing.

So if the lowest half of the income range doesn't pay taxes, and the top rate stays the same, who's getting what changes?  That's what the arguments are about.  From what I hear, the big proposal is a return to three brackets, 12%, 28%, and that 39.6%.  Exactly where those three brackets change will make all the difference in "who pays what".  The simplification part of the plan (they still keep using the "fill out your taxes on a postcard" meme) raises the standard deductions so the lowest income people are expected to pay even less than now, while the highest grade may benefit a little.

The big deal is dropping the corporate tax rate from 35% to 20%.  That will make investing in the US look much better to companies looking to build new facilities.  Lower corporate taxes help everyone but this will have its greatest impact on small businesses and business start ups.  Since it uses the word "corporate", you can be sure the core Evil party hates that part.  Since they can't argue it's a tax cut for the richest, they'll fight that corporations are getting a break.  Corporations pass on their taxes to their customers, so their customers pay them.  Before this all started, Paul Ryan actually said something intelligent, when he said,
"We are actually unique in the world in the way we discourage capital from coming back to America and how we incentivize off-shoring jobs," ... "This is not the kind of exceptionalism we should aspire to ..." 
The overall progressive structure of the tax code doesn't change.  We still penalize success.  We have "sin taxes" on booze and tobacco to discourage their use.  And we have our highest tax rate on our highest earners to discourage success. 
Income and federal income tax, by standing in the income distribution, 2014 bar chart.  The leftmost pair of bars is the bottom 50% of wages.  The blue indicates 11.27% of total adjusted gross income (AGI) was paid to this group and they paid 2.75% of income tax.  The % of AGI is always greater than the percent of the tax burden until you reach the rightmost two.  The uppermost 5% to 1% of incomes group is the first one where the red bar is taller than the blue.  This group earned 15.38% of AGI and paid 20.48% of income taxes.  The top 1% earned 20.58% of AGI but paid 39.48% of income tax. 


Wednesday, September 27, 2017

Why I'm Not Boycotting the NFL

There has been almost a universal "Boycott the NFL" meme rippling through the blogosphere.  Many of you have said it.  I haven't.  I think a little explanation might make some sense.

To begin with, in terms of spending money on football, last year I spent a grand total of $0.00 - zip, nada, nothing.  I think that same sum holds for the previous decade or so.  I actually spent money on the NFL HD video package on DirecTV some years ago, and that would be all I spent.  A friend told me, "viewers = dollars" or eyeballs = dollars if you're online.  To be meaningful to the sellers, I'd have to be a viewer of their commercials.  I don't watch the commercials.  I couldn't even tell you who the advertisers are, except I'm sure there are beer commercials involved from somebody.  I'm not buying anything in response to any commercials during a game.

That means that in real dollars and cents, and in truthful "TV audience", I don't show up in the NFL's ledger at all.  I can't boycott what I'm not supporting with my money.  Do I end up paying tax money to support them?  If I lived in a city or county with a professional team, I'd assume so, but I don't live there.  Does the state tax us?  I honestly don't know, but I'm sure that nothing I could do in front of the TV would affect that. 

Second, I agree that the protests are stupid and juvenile; in other words they're just the thing you'd expect from a bunch of spoiled "man-child" brats.  Protesting "police violence"?  Please.  That's like "hands up, don't shoot", or the picture everyone has of Trayvon Martin as a 12 or 13 year old, not the large, functionally-adult man that died that night.  Both of those are total fabrications, lies from the get go. and to borrow the quote, the lies went around the world while the truth was putting on its pants.  The whole protest movement is based on a lie. 

To hear pro football players talk about being a disadvantaged class because of their race is insulting.  The minimum wage for a rookie drafted in the last round this year is about $485,000 year (average).  For a single person, that's over 40 times the poverty limit in this country, and in the top few percent of all wage earners.  Disadvantaged?  These guys have been pampered all their lives because they can run fast, throw well, or tackle hard.  Even the guys who technically didn't make the team, the practice squad, get $115,200 for the 16 game season.  

There's something more going on here, it's not just this protest stuff.  There seems to be an effort to shut down the NFL and football at all levels.  Think about the fuss about concussions, CTE and players dying from brain injuries.  What about boxers?  Boxers die from getting hit in the head repeatedly.  What about soccer or rugby or any other sport that involves getting hit in the head.  They don't seem to care about those guys, but they're trying to get parents to tell their kids not to play football and only football.  I think that the whole "kneel for the pledge" act is theater; something that Colin Pumpernick came up with to impress his girlfriend.  It's just been incorporated into a larger hard-left plot to get rid of football.  Rush Limbaugh has been talking about this for easily the last two years, if not more.

If it's really a hard-left plot to end football, then, well, why?  As I was putting together ideas on this topic, I ran across this piece on Hot Air, "The Left's (Brilliant) Scam Behind the NFL Anthem Protests."
Why? Because the activist Left has despised the NFL for years. They hate everything about it. It’s a game filled with big, tough, manly men engaging in the closest thing to warfare you can manage without guns. It’s a game rife with symbolism and, yes… nationalism. Even people who would never buy an album from a country singer could feel their blood heating up when Hank Williams used to sing, Are you Ready for Some Football. The military loves football and they fly jets over the stadiums in formation and send our nation’s finest out to pay tribute. And it’s not just the military. Our police and other first responders are frequently called out for honors at the games. Everything about it screams of apple pie, fireworks and patriotism. (Or, if you prefer, God, guns and flags.) And the activist Left hates it. Probably the only sporting event they despise more is NASCAR.
If I watch football, it's for escapism.  I want to avoid politics, and to see really good athletes do things very few people can do.  Avoiding the annoying politics is very easy.  I record the game and start watching an hour or two after the start.  I fast forward until the kickoff and miss the protest.  If the announcers talk about it, I either mute the audio or fast forward a little bit.  I fast forward through all the commercials, too, so I never really see any of them. 

Pro football players are skilled dancing monkeys.  All I want to see is some good dancing. 
Not a hard image to find. 

The inevitable conclusion to this line of thinking is that if we do bankrupt the NFL in five years by boycotting every business that sponsors them, we're handing the left a win.  If that matters.


Tuesday, September 26, 2017

The Feds Issue Autonomous Car Regulations

Self-driving or automated-driving-system (ADS) cars are a regular topic that has consumed a lot of bits here.  (All of my bits are 100% recycled; no electrons are harmed during the composition, posting or reading of a post - although I suspect some would complain of being inconvenienced if they could.)  For ADS cars to really take off as a product, they would need to be one part of an overall integrated system.  The cars would need to talk with each other, which means some sort of digital protocol has to be settled on, radio spectrum would have to be allocated by the Fed.gov, infrastructure for communications will have to be built out, and more. 

On September 6th, the House of Representatives passed the SELF DRIVE act.  As laws these days tend to be named, SELF DRIVE is an acronym: Safely Ensuring Lives Future Deployment and Research In Vehicle Evolution  The act sets up a framework that the industry thinks is necessary
SELF DRIVE ... will allow companies to test up to 100,000 autonomous vehicles on regular roads, even if they don't meet current safety standards. What this legislation does is give companies researching and actively working on self-driving tech a chance to prove their technology is safe.
Machine Design opines:
The need for these regulations is important so that all car manufacturers work against set standards. This way Tesla cannot make ADS vehicles that operate different from Ford, which could lead to accidents. In addition, in the absence of federal regulation, states have taken to creating their own laws for ADS vehicles. This could prevent self-driving cars from crossing state lines.
Personally, I think this is a stupid argument.  Why should industry need Washington to mandate that Tesla not make ADS vehicles that operate differently from Ford?  It's like agreeing on specifications for modems: it's in everyone's best interest to come up with a set of standards everyone can live with; after all, look how successful electronics industry standards like RS-232 or Ethernet or about a zillion others work.  In the world I used to work in, committees of representatives from all of the competitors would meet in "impartial NGOs" to iron out how aviation systems should work.  In the auto industry, I don't know how it came to be, but having the same basic controls - the same basic User Interface - in every car has to help everyone.  Now, could Tesla or Ford or someone else come up with a vehicle that meets all the industry group standard but has some other advantages to try to capture sales?  Why not?  What's wrong with a company trying for competitive advantage by offering something above and beyond the minimum requirements?  I don't see why they want the heavy, clumsy hand of government.

The second part of that paragraph says states might impose regulations such that ADS cars could be illegal in some states while legal in others.  That's (as I've heard it) how the Fed.gov got involved with the unspeakable evil of low-flow toilets.  States had started specifying them on their own and there was no agreement in required gallons per flush.  The toilet companies begged the Feds for "one rule to flush them all" so to speak. 
The second part of the act allows car manufacturers to deploy 100,000 self-driving cars per year (25,000 for the first year the law would be enacted) that would be exempt from existing automobile safety standards. This is important for car manufacturers to design without needing certain safety features like a steering wheel or pedals, such as that for Google’s self-driving vehicles. The act is moving on to the Senate, where autonomous trucks might be added to the bill as well.
The US Department of Transportation issued some guidelines for ADS development and testing.  They eased requirements for safety equipment, so that these test cars are not legally required to have a steering wheel (I didn't know that was a law; I thought we needed one to make the car work). 
At the gathering, Secretary Chao introduced the guidance document A Vision for Safety 2.0, released by the U.S. Department of Transportation and the National Highway Traffic Safety Administration (NHTSA). The guidance, according to Chao, “supports further development of this important new technology, which has the potential to change the way we travel and how we deliver goods and services.” According to the press release from NHTSA, the voluntary guidance will help car manufacturers with the following:
  • Focus on SAE International Levels of Automation 3-5 – Automated Driving Systems (ADSs) – Conditional, High, and Full Automation.
  • Clarify the guidance process, and that entities need not wait to test or deploy their ADSs.
  • Revise unnecessary design elements from the safety self-assessment.
  • Align Federal Guidance with the latest developments and industry terminology.
  • Clarify federal and state roles going forward.
The guidance document outlines 12 safety elements that car manufacturers can follow to help design ADS vehicles. Some of the 12 safety elements highlighted by NHTSA are human-machine interfaces, crashworthiness, post-crash ADS behavior, data recording, and object/event detection and response. Click here for a PDF of the full guideline document.
 (DOT Secretary Elaine Chao presenting at the kickoff meeting) 
 
The "Vision 2.0" safety guidance scales back some of the regulations, reducing the previous guidelines under the Obama administration from a required 15-point safety assessment test prior to deployment to a voluntary 12-point assessment. 

Machine Design looks at this with the view they entitled their article, "One Step Forward, One Step Back".  Why?  They like that SELF DRIVE passed and was sent to the Senate, so they see Federal involvement, but they don't like that the Fed.gov made the safety guidelines voluntary.  They don't like the "light regulatory touch". 

As we've all commented here before, I'm surprised this hasn't turned into a Celebrity Ambulance Chaser Free For All, with lawyers lining up to represent the first next person killed by a self-driving car.  The lawyers are drawn to "deep pockets" like a shark to blood in the water.  There aren't many pockets deeper than the major automakers. 


Monday, September 25, 2017

Decisions, Decisions

After living with my CNC G0704 since last spring, I'm pretty happy with it.  I can see a few things to improve it and reasons not to do anything.  Lately I've been asking myself what's worth doing.  I thought I'd update on what's going on in the shop. 

The three things that would make life a bit better are - in no order at all:
  • Improve the enclosure
  • Implement a standard way of handling and changing tools, probably the Tormach Tooling System
  • Swap the motor - or change the way it's implemented - to double or triple the spindle RPMs.  Also, add turning it on or off and setting RPMs from the computer. 
The short story on the enclosure is that it wasn't very well documented and it's almost a bit of an afterthought.  It's not very robust and isn't the way I think I'd do one if I were starting from scratch.  I've seen better looking, and doubtlessly more expensive enclosures since then. 

The enclosure was based on one that Hoss built for his mill, and like many folks, my conversion was based on his DVD.  Of the many CNC-converted Grizzly G0704 mill/drills out there, it seems that the majority use some version of his approach.  It's not really shown on the DVD I bought, but rather there's a handful of videos on YouTube and many details get talked about in his CNCZone Forum.  

On the other hand, the purpose of the enclosure is to keep the chips and overspray from my misting cooling system inside, and it definitely does that and does it pretty well.  With the exception of the top rail being low so that when I lean inside I bang my head on it about 25% of the time (I am learning to reflexively bend over), it works.  It rattles if you bang it, but the latches hold it down, so this is a pretty minor "nice to have" upgrade. 

The Tormach Tooling System would be an absolutely necessary addition if I was a professional shop trying to recover my costs of doing things.  But I'm not.  For those folks not familiar, the way tools are mounted in mill like this (like most) is with a tool holder that goes inside the spindle.  That holder will hold the cutter, a drill chuck, or other tools.  The G0704 uses an industry standard spindle size, called R8, and R8 collets for all sorts of tool sizes, dedicated tool holders like drill chucks or other things can be found.  The way I change tools now frequently requires me to change the collet and the cutter, then find the new position of the tool.  Back in June, I told about a "touch plate" that uses a piece of software to set the Z = 0.000 inch point. 

What Tormach does is use one slightly-modified R8 collet that is permanently mounted in the spindle.  This is sized for a 3/4" diameter tool, and they have a series of holders that are a 3/4 bar on one end and a custom holder of some sort on the other end.  The feature here is that if the collet is slightly loosened, the tool and holder combination can be swapped with another combination and the system registers the position of the new cutter very precisely in all three dimensions. 
The holders are intended to be used with an automatic tool changer, but can be mounted by hand.  The R8 holder system uses a large bolt (7/16-20) that draws the collet up and tightens the hold on the tool; that task is frequently automated with a power draw bar that releases one tool and then grabs its replacement once the tool is mounted.  The act of tightening the draw bar is what precisely positions the cutter.

In principle, once you find zero for your axes, you always know where X=0 and Y=0 are, and the only axis that might change is just where Z=0 is, if the tool is longer or shorter.  My touch plate can give me that.  What this means is that it might take me extra time to change the collet, change the cutter or drill bit and re-zero, but if I'm not trying to minimize my cycle time in order to keep my price down, do I care?   Except for the obvious neatness factors, I'm not sure I do.

The big drawback to having a separate holder for every tool is that I'd need a separate holder for every tool.  Right now, my tool box has a lot of different size cutters in it.  Many of them need the same size holder, 3/8", but it would be extreme to get one for every cutter in the drawer.  There's a small number that I use the most often.  They would get a separate holder and a permanent place in a software tool table, so that my controller software knows the diameter and length of the tool. 

How much does all this neatness cost?  It could easily run several hundred bucks.  I haven't actually run up a total cost, but I've seen an eBay seller with "clones" of the TTS holders at about $18 each, including shipping (from guess where), if I buy 10.  In addition to those, I'll need collets for the tool end of the tool holder.  There's a few series of standard collets, called ER and then two digits to designate the series.  ER-16, ER-20 and ER-32 are commonly used and I'm likely to need a bunch of those.  In the case of the 3/8 shank end mills prices on Little Machine Shop tell me I'd need an ER-16-9 collet at $9.66 (based on being one of a set of six different sizes), and a TTS ER collet holder at $34.75.  That's $44.41 to hold one tool - and those are "real" TTS tools, not the cheap clones.  As a WAG, I could cut that to about $25 each.  Still not insignificant if I have 10 or 15 tool holders. 

Yet it seems all it really does for me is to make getting a precise Z-axis position easier, a task I've already automated.  Since I'm not trying to shave seconds of a mass production product, it just doesn't seem to matter. 

Finally, the motor.  Here I know the least amount of hard details, but the issue I'm trying to fix is that the max spindle RPM I can get is around 2200.  Another thing to upgrade is that I currently turn on the motor manually by pushing a button on the mill and set the speed with a knob next to it, but both of those operations are simple, standard CNC code.  I just don't have a control that works over the interface yet. 

Virtually every time I'm cutting aluminum, my speeds and feeds calculator tells me to run higher RPMs than 2200.  What's wrong with going 2200 when the software would rather have me go 3600 or 4400?  It wears out the tools faster.  Tool life is a balance of how fast it rotates and how fast it cuts into new metal.  On the other hand, whenever I cut steel, the recommended spindle RPM drops down well below the 2200 limit.  The current motor is specified at 1 HP (750W) and while I'm not sure it even does that (I know... running downhill with the wind at its back), I don't think I'd want to go much above that.  I understand that 3 phase motors are quieter but that will require an electronic box called a Variable Frequency Drive.  A motor like this has been recommended to me.  Whatever motor I'd choose would end up needing an encoder to send speed information back to the controller program, and would need to accept control inputs. 
(Automation Direct GS3-21P0 VFD)

Again, I'm looking at around $500 to implement a better motor.  I might only need the higher RPMs sometimes, but it would be quieter all the time, and I could control it with my CNC control box. 

Since I'm not Elmer J Fudd, millionaire, this raises the usual questions everyone has about the budget only being so big.  Of the three, the motor upgrade might well be the most expensive, but it seems to  make the most sense. 


Sunday, September 24, 2017

The Day Got Away From Me

Didn't pay a moment's worth of attention to the world outside or whatever the Trump Derangement Syndrome Outrage of the Day is, so cartoon!  This is one of my all time favorite Dilbert cartoons; perhaps my most favorite.  I had a faded, slightly-yellowed newspaper version of this on my bulletin board for nearly 18 years back when I was working.  Not quite from the first time it appeared in 1994, but close.
Recently appeared here


Saturday, September 23, 2017

Mythbusting - Economatrix Style

A couple of days ago, The Vulgar Curmudgeon posted a piece called "This is Why You Never Get Ahead".  It's a good economatrix summary, and although I know we've discussed lots of that here, it's good to see lots of good graphics.  Both Raconteur Report (a new link in my blog reading list) and Bayou Renaissance Man did summaries of it.

So why am I here?  It really comes down to one comment from someone who's just not awake to what's really going on; a comment at both Bayou RM and Vulgar Curmudgeon.  I don't want to pick on just this guy because other comments say basically the same thing.
Looking at that chart, we see that a new car doubled in price between 1990 and 2014. That's 24 years. Using the "rule of 72" that I learned in Econ 101, we can easily determine the average annual rate of inflation is 3%. This is also the official rate. I don't see the problem.
Instead of arguing over whether real inflation is his 3% or closer to 9% like Shadowstats calculates (using the methods from 1980 instead of today's methods), I have a different approach.  If you like the 1990 method better, Shadowstats provides that on the same link - using the 1990 method, inflation is closer to 5-1/2%, but again, even though either one of those numbers obliterates his argument, I don't want to go there.  Where I want to go with this is to step back and ask, why should there be inflation at all?  Why do we expect inflation?  The Fed targets 2% inflation, what's a reasonable amount? 

I maintain the ideal value for inflation is 0%.

Since we started out talking about car prices, why should a car get more expensive at any fixed rate?  I can see cars getting more expensive as car makers add mandated features and other costs, but an iron law of manufacturing is that the more experienced you are at making something, the cheaper it gets.  Why aren't cars getting cheaper over the years instead of more expensive?  Are they paying more for materials? Probably, but again, why?  Why should the prices of aluminum, plastic, or iron have anything to do with the nominal inflation rate?   I can see there might be real supply/demand unbalances that make some materials more expensive, but why would supply/demand unbalances produce a curve that looks like "...the average annual rate of inflation?"

Let's not talk about cars.  It's really easier to see with a house.  Have you ever asked why should your house get more valuable with time?  If you're not continually improving it by upgrading things that you'll sell with it, your house should be worth less as it ages, not more.  It's getting older; it's deteriorating.  If the area where you live becomes more desirable, the price would go up, and if your neighborhood has become less desirable, the price would go down. Yet everyone expects to sell their house for more than they paid.  Why should that be?  

Notice the "50 Year Trend Line" of housing prices going up?  Notice that the continuous rise of that line starts around 1971 when Nixon got us off the gold standard? 

Has it occurred to you that our system might be based on a continually devaluing currency?  That's exactly what's going on.  The currency gets worth less and less every year, due to Federal Reserve manipulations.  It's what they're trying to do.  In 1970, gold was $35/ounce while today it's hovering near $1300/oz.  If gold is your measuring stick, today's dollar buys $35/$1300 of what it did then, just under 2.7%.   

There's a difference between inflation and actual economic growth.  Economic growth comes from creating new wealth; either producing things others value out of raw materials, growing crops or mining new things out of the ground: whether minerals like diamonds or metals like iron.  I like to think of this from the viewpoint of the Information Theory of Money: wealth is created by adding information.  A car is good example here.  Creating the car creates wealth for the people who design and make the car.  Now smash that car into a wall at a high rate of speed.  Every single molecule that was present before the crash is present after the crash; what's missing is the ordered information that arranged those molecules into the car.  The pile of parts is worth much less than the shiny new car because the information is now gone.  That's saying that the real wealth; the real worth of the car is the information.  Creating the car consisted of imposing new information on raw materials.  That was the creation of wealth.  Creating currency by manipulating digits in a computer is not creating wealth at all.  What it's doing is diluting all the other currency that already exists, making each unit worth less.  That's inflation. 

The monetarists at the Federal Reserve define GDP growth as economic activity, deliberately conflating inflation with actual economic growth. 

Wages have been in stagnation since the mid 1960s.
On average, workers born in 1942 earned as much or more over their careers as workers born in any year since, according to this research — and workers on the job today shouldn’t expect to catch up with their predecessors in their remaining years of employment.
Workers born in 1942 were probably working for a living by 1960 to 1962, perhaps 1964.  From the group of men that entered the labor market in 1967 to the group that entered in 1983, median lifetime income of men declined by 10%–19%. This problem has been going on for more than the 50 years back to 1967.   Most people will say it's because of off-shoring or outsourcing jobs or blame it on the Evil Rich People.  Bill Bonner had a good summary in a piece I quoted more than a year ago
Most economists (and politicians) have blamed world trade for stagnant U.S. wages. The median wage in China is only $8 a day. No wonder U.S. factory hands can’t catch a break; who can compete with that? 

But Germans compete with the Chinese, too. And their wages have gone up! In real terms, after adjusting for inflation, wages in France and Germany have been going up at a 0.7% rate for the past 15-20 years.
While the Germans and the French have a central bank, they're not working under the US Federal Reserve.  Not having the so-called "Reserve Currency", perhaps their banks are required to be not quite as manipulative as ours, allowing some sanity.  

Well, I often say the only privilege that comes with running a blog is getting to respond with a wall of text to a comment.  Today, I respond with a wall of text to a comment that wasn't even to my blog.  It just seem to me that not enough people ever ask why there should be a constant increase in the prices of everything.
Cars with most of their information mostly removed, rendering them worth far less than when their information is intact. .

Friday, September 22, 2017

US Navy to Use X-Box Controllers on Most Advanced Subs

It's safe to say that only a very tiny portion of the population has actually been in a modern submarine so what we think we know is heavily influenced by Hollywood.  Depending on your age, your first thought is probably either something with tons of analog gauges and dials from vintage movies, or something with computer monitors from a more recent flick.  The fast attack submarine USS John Warner is called "the Most Advanced" submarine in the fleet, and is filled with sophisticated computers with flat-screen monitors to replace much of the conventional instruments and gauges. In those old submarines movies, the periscope was a single optical tube that would be raised above the surface so that a submerged submarine could get a look around.  One man at a time could look.  In the John Warner, those days are gone:
It's been replaced with two photonics masts that rotate 360 degrees. They feature high-resolution cameras whose images are displayed on large monitors that everyone in the control room can see. There's no barrel to peer through anymore; everything is controlled with a helicopter-style stick. But that stick isn't so popular.
The stick is a system designed by defense giant Lockheed Martin - LockMart, as we call them - and it's a $38,000 box.
"The Navy got together and they asked a bunch of J.O.s and junior guys, 'What can we do to make your life better?' " said Lt. j.g. Kyle Leonard, the USS John Warner's assistant weapons officer, referring to junior officers and sailors. "And one of the things that came out is the controls for the scope. It's kind of clunky in your hand; it's real heavy."
The solution?  A video game controller, Microsoft's X-Box controller, something very similar to the Logitech Rumblepad that I just added to my CNC milling machine.
The Xbox controller is no different than the ones a lot of crew members grew up playing with. Lockheed Martin says the sailors who tested the controller at its lab were intuitively able to figure out how to use it on their own within minutes, compared to hours of training required for the joystick.

The Xbox controller also is significantly cheaper. The company says the photonic mast hand grip and imaging control panel that cost about $38,000 can now be replaced with an Xbox controller that typically costs less than $30.
(Photo source)

The crew, after all, is primarily young men who grew up in a digital world playing video games. What could be more natural than a game controller?
The John Warner at her public commissioning ceremony.  Wikimedia photo.
The Navy says that the system has gone through extensive testing over the past two years and that the Xbox controller will be included as part of the integrated imaging system for Virginia-class subs beginning with the future USS Colorado, which is supposed to be commissioned by November.
Just a nifty little story.  Caught my eye because of the similarity to the hardware I've just started using.


Thursday, September 21, 2017

Ambitious... and Impressive

A shop built 1/3 scale model of a V-10 engine with Electronic Fuel Injection.  And, yes, that is a supercharger on top.


Full build details at the Model Engine Maker forums (where you can look at the 20 pages of text and pictures without being a member).  It took about three years of maker Keith5700's time to design and build this impressive little (125cc) engine.  If I understand it correctly, no CNC was involved.  All pieces made by turning the hand cranks on mill or lathe.  Lots of very precise drafting.  Lots of the ideas involved and the skills needed came from an earlier V8 model that he finished in 2013.  

The guy is an amazing artist in metal.


Wednesday, September 20, 2017

What Happens If Maria Meets Jose?

As most of you know, while Hurricane Maria has made all the news today as almost certainly the strongest hurricane in the modern era to hit Puerto Rico, Tropical Storm Jose was sitting off the mid-Atlantic coast as it has all week, aggravating fishermen while exciting surfers.  A storm only sits in one place due to a lack of steering currents, which means the large scale movement of air masses has halted in that area.  A strong low behind a cold front would fling Jose out to sea.  A strong high would fling it into the coast. 

The model runs on the two storms are a study in contrast.  The model runs out to 5 days for Maria show strong agreement that it continues on a more northerly course, paralleling the coast while staying well offshore at least through the 5th day.  The model runs for Jose look like a cat horked up a hairball or someone threw a loose handful of yarn on the map.  A handful of models bring it onshore the Delmarva peninsula, and another group shoves it opposite that, offshore.  Others send it to New Jersey, Virginia, Pennsylvania or most anywhere. 

With my lame photoediting  skills, I combined the NHC plots for both storms into one graphic.  The scales they use are slightly different, so I made them line up around the Outer Banks, NC. 
Simply, both of these can't continue as they are for a full week.  Something's gonna give.  In these plots, it's helpful to remember the stippling means that the storm can be anywhere in that area.  Maria's 5-day location covers what looks to be roughly 500 miles, consistent with the accuracy of about "100 miles per day" that the old forecasters claim.  Jose, though has an oval of uncertainty that looks to be 800 miles or more across - from onshore anywhere from Virginia to New Hampshire,  to still at sea, a bit east of where it is right now. 

So what happens?  There is an uncommon phenomenon called the Fujiwara Effect, named for a Japanese meteorologist, in which the storms start to circle each other and then spiral into each other, becoming one storm.  At their closest - in this plot - they look to possibly be within a couple of hundred miles. With tropical cyclones, they typically interact within about 850 miles, according to that Wikipedia article.  They are absolutely going to be that close to each other, that is, according to tonight's models.  The Fujiwara interaction happens on occasion in the Atlantic basin, but not often.  We may get to see it.

To me, what hurricanes "decide" to do is like whatever two consenting adults decide.  I don't care if they check into a Motel 6 and merge repeatedly.  As long as they keep their Fujiwara-ing away from anyone else, and don't hurt anyone while they're at it.


Tuesday, September 19, 2017

Tesla's Push for Self-Driving Cars Driving Off Engineers

Tesla CEO Elon Musk has famously been pushing for "auto pilot" and fully autonomous cars, saying in October of 2016 that all new Teslas would have self-driving capabilities.  I've been covering some of the autopilot crashes of those self-driving cars as data comes around.  By that I mean that I don't hang out on Tesla user's groups looking for stories of accidents.  On the other hand, the technology needs to get some light focused on it.  It's not ready for Prime Time and Tesla finally seems to be acknowledging that.

This week, Design News presented an article with a few "fun facts" that indicate US consumers are really skeptical about autonomous cars.  First, a Gartner's Survey shows that 55% of respondents would not ride in an autonomous car.
The Gartner Consumer Trends in Automotive online survey, conducted from April 2017 through May 2017, and polled 1,519 people in the U.S. and Germany, found that 55 percent of respondents will not consider riding in a fully autonomous vehicle, while 71 percent may consider riding in a partially autonomous vehicle.
It turns out people seem to have taken lessons from things like the Equifax data breach - along with so many others - to heart and are concerned about security along with the general category of "technology failures".
"Fear of autonomous vehicles getting confused by unexpected situations, safety concerns around equipment and system failures and vehicle and system security are top concerns around using fully autonomous vehicles," explains Mike Ramsey, research director at Gartner.
Gartner isn't unique in finding this; AAA found similar results.
A new report from AAA reveals that the majority of U.S. drivers seek autonomous technologies in their next vehicle, but they continue to fear the fully self-driving car. Despite the prospect that autonomous vehicles will be safer, more efficient and more convenient than their human-driven counterparts, three-quarters of U.S. drivers report feeling afraid to ride in a self-driving car, and only 10 percent report that they’d actually feel safer sharing the roads with driverless vehicles.
Three-quarters of drivers are afraid to ride in a self-driving car?  That's quite a bit more than the 55% Gartner found, but those numbers are more like each other than not.  There's a lot of anxiousness about the technology.  Still, AAA reports that while the majority are afraid to ride in a fully self-driving vehicle, the survey also found that a bigger majority (59%) wants to have autonomous features in their next vehicle.  This apparent contradiction suggests that drivers are ready to embrace autonomous technology, but they are not yet ready to give up full control.  People are responding with a skeptical "prove to me it works" attitude. 

I wasn't aware of the impact at Tesla itself until the Design News article.  Summarized in Inc. Magazine,
Musk's fervent vision of a driverless world repeatedly clashed with his team's view on the progress of the technology. The dispute led Sterling Anderson, then Autopilot director, to resign just two months after the October announcement. Since his departure, at least 10 other engineers and four top managers -- including Chris Lattner, who had been selected as Anderson's replacement and lasted just six months on the job -- have also left.

This isn't the first time dissent among Tesla engineers has been reported. In July 2016, after a fatal crash killed a driver using the Autopilot function, CNN reported that several employees tried to warn Musk about the reliability of the technology, and that it wasn't ready to be deployed to consumers.
The team working on Tesla's Autopilot feature felt uneasy about marketing the technology as "fully autonomous."  They felt they didn't yet have a product that could "safely and reliably control a car without human intervention."  In other words, they felt like the 55 or 79% of consumers surveyed who didn't want to trust their lives to the self-driving cars. 

This is called engineering ethics.  We're always required to voice our opinions and point out unsafe or improper things.  At some point, if the company doesn't respond, the only ethical thing left is to quit. 
The driver of this Tesla says that when she was parking in front of her gym, it jumped into high acceleration, jumped the curb and hit the wall.  Tesla flatly says "nope", their monitors say she stepped on the gas.  Could be.  Hey - it's not your typical Tesla picture, that's all I was looking for.



Monday, September 18, 2017

Turning Aluminum Cans into an AR Lower

There are few 20 minutes videos that I've watched that haven't had me reaching to see if I could skip over some nonsense.  This one had my complete attention for all 19 minutes.  Farmcraft101 takes five pounds of saved aluminum cans and recycles them by melting and casting an AR-15 lower.  H/T to ENDO for the link.


I've got to say his PPE (personal protection equipment) made me cringe a little, but that's the only thing I can be critical of.  Upper arm-length, heavy, leather gloves combined with shorts and bare legs while pouring molten metal is enough to make me cringe.  The rest of it is great stuff to know.

That said, I have to wonder if the metal would be useful for most things.  When you see things saying they're made from "Aircraft Aluminum" or an alloy like 6061-T6 or 7075, that's a specific composition of alloying elements in specific proportions, and T6 is a specific heat treatment.  If I took a pound of 6061-T6 cutoffs and melted those down, instead of soda cans, I wouldn't end up with 6061-T6.  All metals are like this, really.  Steel, brass, aluminum, titanium or whatever, the properties you see depend on the ingredients (alloy) and how they're treated.  Anyone who has taken the mechanical engineering classes on materials has seen something like this iron/carbon phase diagram.  The different colors code for different microstructures in the steel, the temperatures and concentrations of carbon that lead to their formations.  There are similar curves for aluminum and its main alloying additions - silicon and magnesium in 6061 or zinc and magnesium in 7075, for example.

That said, an AR lower has got to be pretty non-critical.  It's not just that plastic lowers are a thing, and can be bought in any quantity from an handful of companies, there's that guy who made one from HDPE - the plastic used to make kitchen cutting boards.  If a cutting board works, it's probably not a high-stress application. 


Edit June 25, 2019 1055 EDT:  The YouTube video had gone away, but I found a replacement version.




Sunday, September 17, 2017

Something Completely Different

While I don't do it often, I keep seeing folks on my regular reading list posting a little music on weekends, especially Sundays.  I don't do that as a rule, and I'm not planning to, but I came across what I thought to be a remarkable performance by Glen Campbell in a link from Acoustic Guitar magazine.


Acoustic Guitar tells the story:
Glen Campbell first heard The Lone Ranger’s brisk theme song as a kid and vowed to learn it on guitar. Not only did he do just that, Campbell made the theme—an overture from the 1829 opera William Tell, by the Italian composer Gioachino Rossini—one of his signature numbers. He revisited it throughout his career, wowing audiences by playing it with casual ease, sometimes with the guitar on top of his head. [link added - SiG]
I would never describe myself as a Glen Campbell fan and never thought of him as a virtuoso guitarist.  In my mind, Glenn Campbell associates with his late '60s hits: Wichita Lineman, Galveston, By The Time I Get to Phoenix, and that's about all.  Maybe the stereotype of country music: "three chords and the truth".  That said, I find this an amazing performance.  Not just the blazing speed, but the casual, almost carelessness he plays with.  Virtually all guitarists, no matter how many years they've played, spend a lot of time watching their left hand on the neck. Glenn glances there, but looks around a lot, too.  He seems to be talking to people out of the view of the camera.

It brings to mind another cliche' I've heard about music in general.  Practice a thing perfectly a thousand times and you play it well.  Practice it a thousand times a thousand and it plays itself through you.  I think the William Tell overture is playing itself through Glen.


Saturday, September 16, 2017

Farewell Cassini

NASA's probe orbiting Saturn, Cassini, ended its 13 year mission yesterday with a fiery entry into the gas giant's atmosphere.
NASA received its last data transmission from the Cassini spacecraft at 4:55:46 a.m. PDT (7:55:46 a.m. EDT, 1146 GMT) today (Sept. 15), before losing contact with the probe as it hurtled into Saturn's atmosphere. It was a fiery grand finale  for the probe, which spent 13 years orbiting the ringed planet. NASA officials expect that Cassini broke apart about 45 seconds after that final transmission, due to the intense friction and heat generated by the fall.
Properly called the Cassini-Huygens mission, the probe was a joint venture of NASA, the European Space Agency, and the Italian Space Agency.  It was launched in 1997 and arrived at the Saturn system in 2004. In 2005, the Huygens lander dropped onto the surface of Saturn's largest moon, Titan, revealing the hidden world beneath its opaque, orange atmosphere.  The Huygens lander detected lakes and seas of liquid hydrocarbons such as methane, ethane, propane and some heavier compounds on Titan.  Liquid methane boils to the more familiar (here!) gas at −161.49 °C or −258.68 °F. , so that gives you a perspective on just how cold it is there.  The temperature at the landing site was 93.8 K (−179.3 °C; −290.8 °F). 

The Cassini orbiter's initial mission was meant to last until 2008 but was extended twice, stretching the spacecraft's life to 2017.
Artist's image of Cassini as it burns and breaks up during entry into Saturn's atmosphere.

There are some "best of" pictures at MSN Look at those at full size. If you know the mission, many of them will be familiar. 

As an aside, a month ago, I did a piece on Voyagers 1 and 2, the first man-made objects to leave our solar system. I mentioned a PBS special that was going to air on September 23, The Farthest. That was the night we got home from our trip to see the eclipse, and I set the DVR to record it.  It's worth watching if you haven't seen it.  Was it everything I'd have liked them to cover?  No.  It was short of some of the important techie details and long on details about the album the Voyagers carry with recordings of music and voices from Earth.  Too artsy for me.  Like I always say, when the typical commentator talks about diversity, they mean people of different ethnicities or colors.  Real diversity is putting engineers and artists together.  I've met enough of all of them to know I have far more in common with black, Asian or Indian engineers than I do with any dancer I've ever met. 



Friday, September 15, 2017

Nanny State Madness - Put Calories on Grocery Receipts

From the Nanny Madness of the UK, a proposal to put "food quality" scores on grocery receipts. 
Well, a new proposal says the “traffic lights” system – widely adopted on food and drink packaging since 2013 – should be migrated from individual items to entire receipts. Forget the problem of making it through checkout without impulse-buying three bags of bonbons and a copy of Heat! magazine; now it is about the half-price pizza catapulting your basket into “red” territory.  

“Instinctively, it seems like a good idea,” says Ed Morrow, campaigns manager at the Royal Society for Public Health. “If health information is just on the product, it’s easy to ignore, but if you get another reminder at the till you might start to compare receipts, see what you’ve scored each time, then try to do better. Doing things that gamify the experience of shopping can be a good motivator in terms of changing behaviour.” That is if you want your midweek trip to Tesco to be gamified: is this just 21st-century code for the nanny state? “It’s not telling people what to do,” Morrow says. “All it does is provide people with extra information.”
Sheer madness.  Let's start with the knowledge that there's no evidence that forcing restaurants to put calorie counts and nutrition summaries on menus does anything.  In fact, the evidence is that the requirement to post nutrition counts makes work for restaurants and raises prices but consumers don't care.  Now let's extrapolate to the grocery receipt.  This is based on the assumption that what the buyer bought is eaten between grocery store trips as it came from the store, and not prepared into meals combined with other things they already have or buy elsewhere.  The buyer could just as well be feeding friends, donating food, or storing it so that it's eaten over a longer length of time or at some time in the future.  All of the nutritional guidelines are based on consuming a mix of foods in meals and the quality of the meal is based on comparison to standard daily intakes of some amount of food.  There is simply no way to get that from a grocery list, unless everything that person buys is  precooked meals from that store and that's all they ever eat. 

Let's go with a simple example.  Someone buys some food that the food police code as red.  Then they go to another shop and buy other things coded green to make a meal that changes the nutritional data completely.  The article states it's likely to be accurate in the long run, but I don't see how.  Averaging bad data doesn't make it good data.  It just makes a bad average.
The proposed "traffic light" method.

Mrs. Graybeard and I have a running joke that the purpose of buying celery is to put it in the refrigerator until it gets limp and has to be thrown out.  That's an exaggeration, of course (exaggeration is the essence of humor), but we honestly think we've thrown away more celery than we've eaten.  That's a way to bias the data, too.

Nutrition codes on a grocery receipt are meaningless information that aren't going to affect anything, are based on several faulty premises, and will create costs that the poor customers are going to have to pay for - so that they can ignore it.

But remember the line in the first clip from the article where Ed Morrow said, “It’s not telling people what to do?”  Anna Taylor, executive director at the Food Foundation. removes any doubt that telling people what to do is exactly what this is about.  It just uses a Cass Sunstein "Nudge" approach.
“What would be really interesting is if the retailers link it to algorithms set up for loyalty cards for those who want it,” says Taylor. “So, if a till receipt shows lots of reds, you might get vouchers to buy more veg. That’s when it could start to get really powerful. But let’s be clear: the whole food system has to work harder to make the easiest choices the healthy choices. That’s what needs to change.” 
No, Ms. Taylor.  What needs to change is you.  You need to go away. 


Thursday, September 14, 2017

One Last Irma Post

I promise.  It's just that the whole subject kind of has a built-in expiration date.  Denninger has a similar post with some similar data, so I thought I'd show some more. 

During the day Sunday as the storm was working its way north, I recalled watching the wind speeds reported back from NOAA buoys during Matthew and thought I'd watch Irma approach.  There's a webpage with a graphical front end,  that allows you to zoom in on an area, click on a buoy's location, and read the conditions its reporting.  Not every one reports everything, so you may have to experiment to find what you're looking for.  Here's the map you see when zoomed in around Florida.
You'll notice I have three of the yellow diamonds circled.  These are Key West, Vaca Key and Naples.  (Red diamonds are buoys that are offline).  If you'll recall, the storm crawled along the north coast of Cuba, then hung a right and went for the Keys.

By the time I remembered to look up this web site, it had already gone through and I didn't know where it went.  So I looked at both Key West and Vaca Key (on Marathon).  Neither of them had a wind measurement over 60 knots - 69 mph while the official winds were 150.  How could they miss it?  Were they both so far from the eyewall they missed the strong winds?  Now the distance from Key West to Vaca Key is about 50 miles, and Hurricanes 101 says the strongest winds exist only in eye wall, but it's still a puzzling result.  Did the hurricane thread the needle between them? Yes, it turns out it did, making landfall at Cudjoe Key, which is virtually the halfway point.  If the eye is 10 miles diameter (a wild-assed guess), that means about 20 miles from the innermost eyewall to either buoy.  I know the winds tend to drop off with distance, but that fast?

One of the reasons I thought to come to look at these buoys is that I noticed the same thing last year during the buildup to Matthew.  Official winds were in the neighborhood of 150 mph, but I never found a buoy that went above 100, even ones the storm had gone right over.  Unfortunately, I didn't write anything down. 

Still, I was puzzled.  So as the storm crept toward Naples, I picked that buoy to monitor.   Here's a screen capture.
I've highlighted in red the time, between 4:54 and 5:00 PM, that the eye went over the buoy.  The winds dropped to under 10 knots and the barometric pressure dropped to its minimum 27.75 inches.  But note the winds right before and right after the eye passage: 22 knots: 25mph.  Now look for the highest winds near those times.  The highest wind before the eye was 42.9 kts or 49.4 mph.  The strongest wind after the eye passed was 55.9 kts or 64.3 mph, an hour after the depth of the eye.  These are not a hurricane force winds, which start at 74 mph. Irma was not a hurricane when she hit Naples. 

I keep coming across little things that make it seem that they systematically overstate the winds.

For one example, on Saturday morning, one of the admins at Central Florida Hurricane posted that he was watching the wind speed coming in from the hurricane hunter aircraft, and they were showing the wind at 122 mph.  Minutes later, the Hurricane Center posted the official forecast and it said winds were 155. Why?  On the other hand, they dropped the winds to 125 on the next update, so maybe it was just too late for them to incorporate the new data.

For another example, I found this interesting comment on Watts Up With That by a guy calling himself FLengineer (no relation).  There's a lot of detail here on NOAA changing the way they report winds.   The quote is actually too long to post here, but let me grab a little bit of it:
After digging into the specs and extracting the measurements from the transmitted data it looks like the advisory intensity is purely based on a peak wind speed measurement over a 10 second interval at flight level between 2200-3000 m altitude. I found a decent paper on the measurement equipment aboard the aircraft from 2007 (http://www.aoml.noaa.gov/hrd/Uhlhorn/MWR2007Uhlhorn.pdf) and thought some of their conclusions were interesting.

Prior to 2005 intensity was determined by a model that took the flight level winds and extrapolated 10m surface winds from those. In 2005 the planes were outfitted with an updated SFMR radar that measures the ocean surface emissivity which is correlated to surface wind speeds. Due to lots of tropical activity in 2005 the new radar was able to help derive a better physical model matching the readings with dropsonde data.

In their conclusions the earlier quadratic models consistently underestimated winds at speeds > 50 m/s (111 mph). They also found that earlier boundary layer models were underestimating winds in the eyewall by as much as 10% when compared to dropsonde readings.

So are the storms more intense because of increased SST or because of better models, higher frequency sampling, changes to peak measurements from 1 minute averages, and improvements in the sensor equipment?
...
If we look at the most recent Air Force data in Irma the peak flight level wind was 133 kts (153 mph) SFMR surface wind was 120 kts (138 mph) and same 88% linear adjustment would extrapolate 134 mph. NHC 11 pm advisory for Irma is 160 mph.
What does it all mean?  Beats the heck out of me!  Here we have several lines of data showing that when winds are measured in the monster hurricane, it's not what's being described. 

I know that measuring and quantifying these things isn't easy, but there seems to be a systematic tendency of the NHC to over hype every hurricane.  The NHC itself has not been friendly to the "global warming causes hurricanes" crowd, and one of the NHC's chief scientists, Dr. Chris Landsea, withdrew from the IPCC reports over their reports not being backed by science.  Maybe they're just covering their uncertainty about what to forecast by getting everyone in the state to leave.  We know the Weather Channel and virtually all of the news channels make their ratings and their advertising money on storms.  I know that I've seen TV weather folks whom I respect who will tell you to your face that "we have to make it as dramatic as possible to get people to leave".  But what if they don't have to leave?  What if leaving put them in more danger from other things.  How many stories did you hear of people who evacuated south Florida to the west coast and ended up going right into the worst possible place? 

I've seen this described as the biggest mass evacuation in history. 

Evacuating costs real money.  Did they negatively impact someone's life by having them take on the burden of running from the storm?  You've got to know that a lot of people can't really afford to drop whatever motels were asking for a few nights, and even more for meals.  What if they came home to a looted home? 

Let me leave you with a fun fact.  After looking around at various buoys from the keys up through Ft. Myers, and looking at some reports off that NOAA bouy page, I found one buoy with hurricane force winds: Fowey Rocks buoy, off the coast of Miami, pretty much 115 miles from both Cudjoe Key and Naples. 


$318 Billion in One Day

A few weeks ago, I said there's no debt ceiling.  Just ignore the hype coming out of DC and know that you can absolutely count on them not being grownups, not showing any fiscal restraint and eventually raising the debt ceiling.  There has been a legal debt ceiling since 1917 and they've never not raised it before, so why should we expect that with the fractured mess we have in DC now that they'd suddenly grow up now?

As predictably as the coming sunrise, a deal to increase the debt ceiling was reached; the only novelty being that this time the president dealt directly with the Evil Party instead of the traitorous members of his own party.  (As an aside, while we decry Mitch McConnell for being spineless and the Stupid Party for being unable or unwilling to do anything they ran on doing, you have to credit Chuckie Schumer for being, perhaps, the most effective senator on the hill, simply for the amount of time he gets his way).  As they've done before, the Treasury Department magically slowed the rate of debt increase by special means, the rest of us call that "lying", and then on last Friday the national debt suddenly and equally magically increased $318 Billion in One Freaking Day, so that now we're well over the $20 Trillion National Debt level.
“Yeah, we need to cut spending, but we can't do it now!  We need to provide relief for those Hurricane Harvey victims in Texas!  It's different this time, not a regular debt ceiling!”  Don't worry, though, there will be something else next time to prevent the discipline.  Don't we remember Pelosi saying there's nothing left to cut in the budget?  The fact is that they'll never cut the budget.  It just ain't happening.  There are far too many people and groups who depend on that flow of money - and those people and groups funnel money to the ruling class to perpetuate the system.

Over a dozen years ago, Dick Cheney famously said, “Deficits don’t matter.”  In the short term, that might be true.  Seed corn wouldn't matter if you knew there'd never be another tomorrow.  Hungry?  Eat the seed corn.  If there was no tomorrow, there'd be no reason not order a few extra desserts, no reason not to have a few more drinks, no reason not do just about anything.  But there is a tomorrow.  You need to plant the seed corn so you can eat next year.  You might find the extra drinks lead to a hell of a hangover.

Since there is a tomorrow, the debt will matter.  I can't tell you just when or just how it all comes crashing down, but it's fake.  There is no real wealth backing it up, just phony money.  It's the biggest pile of phony money in the history of the world, and it's going to make the biggest mess in the history of the world.