Saturday, February 29, 2020

Offered Without (Much) Comment

Last night, SpaceX destroyed Starship prototype SN1 (serial number 1) during a liquid nitrogen fill test at their Boca Chica facility in Texas.  There had been talk that SN1 was going to take the first suborbital flight for a Starship.  I suppose it did, because it certainly lifted off and never obtained orbit, but I don't think this is what they had in mind for a mission profile. 



I don't have much comment because there has been no official story on this, and this video comes from the forums on NASA Spaceflight.  That link will take you to the end of the thread, which goes back to last summer.  Currently that's page 89 and the accident seems to have happened during page 87.  There are interesting pictures of the wreckage taken this morning in daylight. 

Just three days ago, on the 26th, Teslarati was excited that SN1 had been rolled to the test stand and mounted.  There is no coverage of this accident, but perhaps the fact that it happened on Friday night means the staff was off and they don't work weekends.


The forum at NASA Spaceflight has idle speculation on what the cause of the problem could be, but until some sort of investigation is completed, that's just speculation.   First law of breaking news applies here: the first reports are always wrong.



In totally unrelated SpaceX news, Monday's planned Falcon9 launch of a resupply cargo capsule for the ISS has been delayed until the Friday, the 6th.  The reason is pretty remarkable and the only reason a launch delay is worth mentioning.

During tests of the system that was considered ready to launch, a problem was found in the second stage.  Thanks to their aggressive launch cadence, they have plenty of hardware and chose to just replace the second stage with another that was scheduled for a Starlink internet satellite launch.
“SpaceX identified a valve motor on the second stage engine behaving not as expected and determined the safest and most expedient path to launch is to utilize the next second stage in line that was already at the Cape and ready for flight. The new second stage has already completed the same preflight inspections with all hardware behaving as expected. The updated target launch date provides the time required to complete preflight integration and final checkouts.”
NASA.gov — February 25th, 2020
Given the complexity of an orbital rocket's upper stage, it's a safe bet it takes a couple of months to build one and get it ready for flight.  To replace an upper stage and just delay a launch a few days can only mean they basically had one there on the Cape ready to fly.  NASA gets what's likely the safest scenario, replace the upper stage entirely, and SpaceX can troubleshoot this upper stage as they feel is best.


Friday, February 28, 2020

Space Industry First - A Satellite Docks With Another

In an historic first, two unmanned satellites docked with each other in orbit this Tuesday.  The feat was announced by Northrup Grumman, whose SpaceLogistics subsidiary created the satellite that achieved the docking.  Northrup heralded the mission as an "historic accomplishment" in the field of satellite servicing. Prior to this mission, no two commercial spacecraft had ever docked in orbit before.
Northrop Grumman Corporation (NYSE: NOC) and the company’s wholly-owned subsidiary, SpaceLogistics LLC, have successfully completed the first docking of the Mission Extension Vehicle-1 (MEV-1) to the Intelsat 901 (IS-901) spacecraft in order to provide life-extension services. This historic accomplishment marks the first time two commercial satellites have docked in orbit and the first time that mission extension services will be offered to a satellite in geosynchronous orbit.
MEV-1 was launched from Russia aboard a Proton rocket last October 9th, and the satellite's small ion thrusters took three months to get the satellite high enough to reach the geostationary orbit belt at about 22,250 miles, where the orbital period matches the Earth's rotation and the satellite stays motionless in the sky as seen from the ground.  The Intelsat (IS-901) spacecraft was launched in 2001 and was pulled from active service in December 2019 as it ran low on fuel.  Operators commanded the satellite to move into a "graveyard orbit" farther out than the unique geostationary space. It is here that MEV-1 linked up with the communications satellite on Tuesday.  Northrup uses a mechanical docking system that can be glimpsed in this animated GIF from the MEV-1 as it captures the IS-901.  There's a short graphic at Popular Mechanics that might help you visualize it better.  The system latches onto existing features on the satellite; in this case the satellite's engine.  Northrup says it's designed for multiple docking and undockings and can deliver over 15 years of life-extension services.


This is not refueling on orbit, as we talked about last August; the Intelsat IS-901 wasn't designed to allow refueling.  I'd be sure that's true for everything else currently on orbit.  The Mission Extension of the MEV-1's name is achieved by grappling the satellite and having the MEV-1 bring it back into the geosynchronous orbit where they will remain attached for years.  Ars Technica reports:
According to Northrop Grumman, the combined spacecraft stack will now perform on-orbit checkouts before MEV-1 starts to relocate the combined vehicle back into geostationary orbit, where Intelsat 901 will continue in service for five additional years.
Northrup plans to launch a Mission Extension Vehicle-2 later this year.  Northrop also said this is its first step toward establishing a fleet of satellite servicing vehicles that not only extend the life of satellites like this mission, but provide orbital inclination changes, can provide spacecraft inspections and perform in-orbit repair and assembly. 

Final words to Eric Berger at Ars Technica:
As Earth orbit becomes more congested, a number of companies are working to develop technologies to both service vehicles as well as remove orbital debris from space. Tuesday's successful docking seems to be a positive step forward into a future in which on-orbit spaceflight becomes more sustainable, where spacecraft connect, assemble, and operate in a more coordinated way.

Wednesday, February 26, 2020

Living in the Future - Memes From Space

I'm a bit surprised to see by checking my blog history that I bought my first RTL-SDR in November of 2013, so 4 months more than 6 years ago.  If that's nonsense language to you, SDR is short for Software Defined Radio, and the RTL-SDR (I have this one) is a particularly cheap radio that receives over a wide frequency range, 24MHz to 1850MHz, by hacking the software in a chip (RTL2832 - where the "RTL" in the name comes from) designed for receiving HDTV over the air.  With an open source software package, SDR# (SDR Sharp) in my case for a desktop PC running Windows, it will demodulate virtually everything you'll hear in those bands (I'm not aware of anything it won't demodulate).  It's simply the cheapest broadband radio you're going to find. 

There's a website devoted to all things open source and hobbyist SDRs called, appropriately enough, RTL-SDR.com/ that I've started following more than usual in the last few months. There's always something interesting going on there and it's updated every couple of days.  It was here that I came across an open source project that could only come from living in the future.  A small group is working on an open source satellite; a Cubesat, the form becoming common in education and frequently put into orbit either free or at bargain basement rates.  The sole purpose of this satellite will be to transmit memes back down to Earth.
The Mission for Education and Multimedia Engagement Satellite (MEMESat-1) is planned to be the first meme broadcasting cube satellite ever created. If you aren't down with modern slang and are not familiar with the word "meme", that may be because although first coined in 1976, the modern definition was only added to the Webster-Miriam dictionary in 2015. In the traditional sense a meme is a cultural idea, behavior, style that people can't help but want to share because of how funny/amusing/interesting it is.

But in particular MEMESat-1 wants to broadcast from space the new type of meme definition, which is essentially funny or amusing images/GIFs that internet users and especially youth like to modify and share online through social media. Memes have become a major part of internet youth culture, so this could be an excellent way to speak the language of the next generation and get them interested in space, satellites, amateur radio and building satellite ground stations.

At the moment, the team hopes to launch the satellite by late 2021, and no later than Spring 2022. The satellite will be a cubesat with flash memory containing thousands of meme images that will be broadcast to Earth via a transmitter operating in the UHF 70cm radio band.  Enthusiasts on the ground will be able to receive the meme images with a Yagi antenna and we anticipate that RTL-SDRs will be a commonly used receiver. The satellite will also contain an FM UHF/VHF repeater operating in the amateur radio band for ham radio use.
The project has its own funding website, Let's Go To Space, where they're trying to raise $30,000 to build and launch the satellite.  Donations of any size are welcome, and if you're a Meme-Generating God (I'm thinking of you, Aesop), you can get your memes into the memory on the satellite to be cycled down to earth for $1.69 each.  Whenever the total donated goes over a thousand dollar threshold, memes will be free to upload for 12 hours.


On the ham radio side of that, the 70cm band is from 430-450 MHz in the US, and when they refer to “an FM UHF/VHF repeater,” the existing amateur satellites generally uplink and downlink on different bands; with some using VHF (2m - 144 to 148 MHz) up and UHF (70 cm) down but they're not all that way.  Certainly, if you had one of the ubiquitous Chinesium HTs like the Baofeng or similar and something like one of these Arrow antennas, you'd be able to hear the memes, but you'll need to interface the recovered audio to your computer to turn the tones into the picture.  I don't want to get too much into details on working the satellites here - I've never done it. 

This is detracting from the story, though, which is fun to read and you should go RTWT.  The original story is on Reddit, but is excerpted on RTL-SDR.com.  It's the story of a Grad student in his early-20s who started out joking with friends about this, but happens to be in a circle of friends that can do something about turning those jokes into reality.  Such that he now has backing from industry giant Ball Aerospace and Blue Canyon Technologies.

Another “what a time to be alive” story.


EDIT 02/26 2229 EST:  Typo in the first paragraph on the dates.  Thanks to comment from Chuck Pergiel.


Tuesday, February 25, 2020

The US Fed.Gov Sells Cigarettes?

Yeah, but for once there's actually a good reason.  This is in today's Machine Design newsletters, reported by Steven Mraz.  Chances are it's still contracted out to a company that knows how to do it rather than being manufactured by the Fed.gov.

The cigarettes are sold by the National Institute of Standards and Technology (NIST) and are called NIST’s Standard Cigarette for Ignition Resistance Testing, or Standard Reference Material (SRM) 1196.  They're used in testing mattresses, furniture and other furnishing as way to measure flammability.  Apparently, they're used by lighting one up (a smoking robot? surely not a person hired to smoke in the 21st century!) and putting it on the mattress, or curtain or whatever to assess how flammable the consumer item is.  


Bender the robot on break from his job lighting SRM 1196 standard cigarettes.  (No, I'm not a Futurama fan)
 
The SRM 1196 cigarettes are non-filtered, longer than commercial cigarettes and are made using unbanded paper. (Banded paper contains sections of less permeable paper that lowers the amount of air available to the burning tobacco. They’re prone to go out on their own unless smoked, so they are less likely to start fires than the SRM 1196 cigarettes.)  The NIST sells them for $1 a piece, in units of 400 (two cartons of 200 cigarettes each).   

We can joke about it, it's hard not to joke, but it's serious stuff.  Despite the drop in the number of smokers in the general population of the US, one out of every 20 home fires is caused by a cigarette, pipe or other smoking material, and one out of every 31 home fires causes a death.  

The idea that a standard ignition source is the way to test things is probably something that everyone reading this with an industrial test background is nodding in agreement with, and people who don't come from the test or quality control backgrounds are puzzling over.  Since the SRM 1196 became standard, more and more organizations and industries adopted regulations that called for the use of the reference cigarette. These include the American Society for Testing and Materials (ASTM), the National Fire Protection Assoc. (NFPA) and the Consumer Product Safety Commission (CPSC).  As a result, although the NIST ordered what they thought was 10 year supply of the cigarettes in 2010, they ran out two years ago.  In response, they ordered a new batch, but called these SRM 1196A despite supposedly being the same formulation. 

As I read about this, I'm struck with the thought that this seems like it should be up the alley of the insurance industry; something like the Underwriter's Laboratory (UL).  I have a hard time thinking the NIST is manufacturing cigarettes; they probably contract with a cigarette company to make them to government specifications.  Surely a private company, funded by the insurance industry, would be just as interested in improving fire safety and would see the usefulness of a standard ignition cigarette.  Let the NIST define the standard if the industry can't agree, perhaps including how to use the standard cigarettes, then have the NIST hand over the data to some group like the ASTM and get out of the business.




Monday, February 24, 2020

Chairman Xi's Comments and the Covid-19 Virus As a Weapon

If you read LL's excellent blog Virtual Mirage, you probably came across this quote in Sunday's "What Will Happen (Sunday Sermonette)" In answer to a comment, LL wrote:
President Xi went on the record, essentially saying that it was an accidental release from their own bug factory yesterday. Some media outlets picked it up. If the White House suggests that it's an engineered product, I think that the DJI will tank and drag a lot of stocks with it.
His perspective is that the economic damages are the most frightening thing about this likely pandemic.  With the field that the Evil Party has put forward, the party is exceptionally weak. I have no faith in polls 9 months ahead of the general election.  In a year that's looking like we could see a presidential sweep along the lines of Reagan's second term, I look at the energy of the two parties.  There's little reason for Stupid Party members to go to the primaries or caucuses, yet Trump's vote total in New Hamster was a record high for an incumbent president.  Everybody is crowing about Bernie's victory in Nevada, but the numbers are pathetic.  Nevada has over 2 million adult residents over 18.  Of those 2,000,000, the early voters measured 76,000, a mere 3.8%.  A bit over 6,000, 0.3%, voted for Bernie.  In New Hampshire, a smaller state, over 120,000 people voted for Trump.  

There seems to be only one thing that could threaten that reelection sweep, and that's if the Kung Flu gets really bad and it appears the CDC has completely screwed the response.  Today, the stock markets started reacting to the cases showing up in Europe (Italy, Austria, Russia) and the Mideast (Iran, Lebanon).  As a result, the major indices were down.  If China stays shut down for more months, their economy will crater and the rest of the world with it.  I'm talking not just recession but '08 levels or even worse. Then you have this swell of people wanting bigger government and wanting government to run healthcare motivated even more by fear of the virus.  (Yes, they could be mad at the CDC for not responding properly but still want the CDC to be bigger and control more!)  If people are desperate for medical care and a wild-haired socialist is promising unlimited free health care, that might attract enough voters to turn things.

On the other hand, as Larry points out, “...when you look at how many actual votes he received in Nevada, the Donkey enthusiasm even for free stuff, is very tepid.”

Today's email from FreePressers linked to an article which contained this reference:
Chinese supreme leader Xi Jinping on Feb. 14 told an emergency meeting in Beijing that a national system to control biosecurity risks was needed “to protect the people’s health.”

Xi called lab security a “national security issue.”

While not actually admitting that the deadly coronavirus that is devastating large swathes of China had escaped from one of the country’s bioresearch labs, “evidence emerged suggesting that this is exactly what happened,” Steven W. Mosher wrote in an analysis for the New York Post on Feb. 22.
That's about 8 days earlier than what LL is talking about, but signals the same idea.


President for Life Xi at the conference where he talked about lab security as a “national security issue.”  Licensed under Wikimedia Commons N/A



Sunday, February 23, 2020

The Continuing Saga of the Webster Piston and Cylinder

A followup to my post from a couple of weeks ago (the 5th).  First, I'll borrow the piston illustration so you don't have to go to that link.


After the excellent comment from Ownerus, I put myself back to school on measuring IDs with the telescoping gauge.  Instead of the digital calipers I was using, I went over to my micrometer. 

Before I start, the drawings say to machine the cylinder to 0.875" Inside Diameter and the piston to 0.873.  The rings are to predictably take up the extra .002"

I measured both ends of the cylinder three times.  For the combustion chamber end of the piston, the first two times I measured .8827 and .8829, a difference of "2 tenths" - .0002".  On my third measurement, I got .8837.  Even averaged that's 0.8831.  That gives me some confidence the combustion chamber is close to that 0.8830.  On the opposite end of the cylinder I measured 0.8827, 0.8812 and 0.8817; more spread and average of 0.8819".  This says the cylinder is a little narrower at the end the piston mounts from, and I suspect it would be better for any taper to be a little narrower at the combustion chamber end. 

The problem was that when I measured across the piston rings, I got 0.896.  Which explains why I couldn't get that piston into a hole .014 smaller than the piston. 

After the comments on that post, I had decided to remove the piston rings and deepen the slots because the rings stood much too high above the 0.879 piston.  The rings didn't feel right and I thought the reason was that I made the grooves too shallow.  Plus, the rings felt too tight in the grooves, hardly movable at all.  Here are the results from trying to remove the rings today.


You can see the portions of the shattered piston rings next to the piston. 

For an engine like this that will be run a few times at relatively gentle settings, a polymer (Viton) piston ring is probably a viable choice.  That means making a new piston to fit the too-big cylinder, and that's probably the best thing to do.  Remember, the drawing had them .002 apart in size.  By that, I should make a new piston 0.880 or 0.881 in diameter.

I've got to say I'm a bit tired of working on these two parts, and want to get on with the engine, but it's worth trying to get this right now.  There's a handful of these critical parts in the engine - the valves in their valve cages are the biggest ones - and the rest is pretty routine.  I hope.



Saturday, February 22, 2020

My First Post That Went “Viral”

As part of this 10th year, I'm going to take a clue from one of my role models, Borepatch, who has posted a few of his memorable posts from his first 10 years (for example). 

From September of 2010, On Germs, Weeds, Companies, Governments and Skunks.  This was the first post I can recall getting a great deal of attention, and it was a heady experience to get linked to by Doug Ross @ Journal.  I got links from more places than I recall and for years this was my most read post. 

At seven months short of 10 years old, I re-read it to see if I still thought it was worth standing by and I think it has held up well.  I still like it.  I find the wording a little awkward in places, but that happens with things I wrote 24 hours before.  Most of the links are broken and some of the references might seem obscure.


On Germs, Weeds, Companies, Governments and Skunks

Has it ever seemed like we, as a country, can't do anything anymore?  The country that built the Empire State building in a blink has a gaping hole where the World Trade Center used to be for 9 years?  The country that went to the moon from nothing in 10 years now takes 20 years just to decide if we want to?  There's a reason for that national arthritis that keeps us from moving: regulation and its bastard sibling litigation.  

Credit Borepatch with this long post.  This post last week got me thinking along these lines.

When I was but a newly hatched larval engineer at a major defense electronics contractor, a mentor systems engineer gave me a reprint of an article from Defense Electronics magazine for August 1979, by Norman Augustine, the CEO of Martin Marietta (now part of Lockheed - most of us call them Lock-Mart these days, your one-stop defense supermarket - watch for the flashing blue light special on cruise missiles…).  The article covered what's now called Augustine's Laws.  I see they are out in book form:  and some of them have been excerpted by Political Calculations blog, but the version I have has a mere 15 laws.

There's a lot of wisdom in here and it's worthwhile to see if the ensuing years have changed things much.  I don't have the space to go into everything, but a few of these are worth looking at.  The first law is "By the time of the tri-centennial, there will be more government workers in the United States than there are workers" based on the growth rate at that time.  I think he's a bit long on this; it'll be before 2076.  He goes on to say that the government will control all of the money in the US economy.  "This raises the interesting question of whether the last person employed in the private sector will have to support the entire nation's work force, or whether he or she will individually enjoy the full benefit of those residual funds not yet controlled by the government".  This raises the guardedly optimistic second law, "People working in the private sector should try to save money if at all feasible.  There remains a possibility it may someday be valuable again".  What a silly optimist: the Fed has had crosshairs and laser sights on the backs of savers for almost a hundred years!  He presents a wonderful comparison of a contract for muskets for the continental army in 1798 that was delivered in 1/3 more time than contracted, and a survey of military contracts from 1978 that were delivered, on average, in 1/3 more time than contracted.  Some things never change!

His most memorable law, and where I'm going with this, was "Systems of Regulations created as a management surrogate take on a life of their own and exhibit a growth history which closely parallels other living entities observed in nature".  He went on to show the number of pages in armed forces procurement regulation vs. time along with a curve of weed growth (from the journal "Weed Science"), and produced a graph any biology student will instantly recognize as the sigmoid growth curve of populations, also called the logistic function. 

A usual example is the common bacteria E. coli.  This species can divide and produce a new generation every 20 minutes; if conditions could remain optimum it would undergo geometric growth and produce a colony the size of the planet in 24 hours.  Because conditions can't remain optimum, it has a logistic growth curve, producing much smaller colonies.
The normalized logistic function.  Credit

In regulations, there is a price for this.  Although the legislators and regulators never consider this, every regulation consumes some amount of time and money to comply with.  The new Finance Reform bill has been estimated to required the development of 250-300 new regulations.  Every regulation slows down, hinders and costs every honest business real money.  Despite the widespread talk of corrupt CEOs and general lack of corporate ethics, I've been working in the manufacturing industry since the mid 1970s, and every company has had an active, if not aggressive, ethics compliance program with requirements for training and seminars every year.  There are exceptions but most companies do their best to be honest and law-abiding.  Government seems to think it's mere coincidence that countries with lower tax rates and lower regulation attract business, and they demonize companies for moving to countries where the environment is better. 

A simple way of determining if someone you talk to has any economic sense is to ask them about corporate taxes.  The economically ignorant (I'll be polite) will scream to tax the corporations.  Those with sense will tell you corporations are fictitious and can't pay tax.  Tax is part of the cost of doing business and therefore passed on to the buyer (the people calling for them to be taxed).  Corporations can collect taxes for the government (for which they are punished with more costs, not paid) but cannot generate them.  Every penny a company has comes from its customers.  In a global market where they compete with companies in cheaper environments, they are at a disadvantage. 

At least in our society, regulation often leads to litigation, and companies are sued regularly for offenses (real or imagined) against the environment or against people; possibly employees, possibly customers, possibly people in the communities they work in.  When special interest groups, like environmentalist or animal rights for example, don't get the laws they want passed, they litigate to prevent companies from doing their business.  For example, some of the most verdant farm land in the country, the central valley of California, is currently not being farmed because a group got the courts to stop the irrigation on the basis that it harms a minnow, despite considerable scientific disagreement that irrigation has anything at all to do with the minnow's population.  (Summary here)  The human impact in terms of unemployment and lack of food production, however, is undeniable. 

Regulation and litigation are sand in the gears of society.  Big, sharp, 40 grain silicon carbide abrasive particles that grind the gears and shafts away.   

They have the awful side effect of discouraging companies to grow and innovate.  Industries can die off because of the threat of litigation, or because the regulations make it simply too expensive for the industry to survive.  Sometimes this is deliberate: we just survived the EPA banning lead ammunition, something that could well kill off hunting and recreational shooting in the country.  The requested regulation was a back-door attempt to do something that can't be passed as a law.   

Sometimes killing an industry is an unforeseen side effect.  The woodworking tool industry is in crisis right now.  A few months ago, a lawsuit was won against Ryobi tools' parent company, for producing a "defective" table saw.  The plaintiff was awarded $1.5 million when he sued for $250,000.  The defect?  It did not include an expensive safety option that was invented around year 2000; (the modern tilt-arbor table saw was invented in 1939; the basic idea goes back to 1813).  In 2000, an inventor produced a technology called a Saw Stop that senses when flesh touches the blade and stops the blade in milliseconds.  In the process of stopping the saw as fast as it does, it destroys the saw blade, and possibly other parts of the saw.  The user still gets cut, but typically will require stitches instead of having a body part cut off.  He shopped this invention to the major tool makers and none of them agreed to license his invention.  Their major concern was that the idea was untested; they had no idea how durable it would be (contractors' tools live a rough life); they had no idea if it could be added to existing products (were they rugged enough to survive the abrupt energy dump that destroys the blade?), or how to roll it out across their product lines.  The inventor started his own company, and sells table saws with this feature.  

This suit will end the production of low-priced and bench top table saws, seriously impacting hobbyist woodworkers as well as the tool industry.  Professionals will buy the more expensive saws and raise their prices to you and me.  
 
I'm most familiar with the electronics industry, so let me give you a story from my world.  Most people have heard about Silicon Valley, which certainly was, and may even still be, the innovation capital of the world.  A typical scenario might be like this.  In satellite communications, there is a couple, of large companies who dominate the industry.  Like all large companies, they have a large set of policy and procedure manuals that show everything that must be done for any situation that comes up.  Perhaps a new market is perceived; the company must conduct a study to see if there's enough money to be made to pursue the market.  They must do a developmental budget, have decision agreement ("buy-in") from many levels of management, and so on.  Perhaps a previous customer had sued them over delivery of a product that they thought didn't meet the contract promises, for example, so they may require marketing surveys to understand just what the market wants. 

Somewhere in that company is an experienced engineer; typically someone with 5-15 years of experience.  They see the opportunity and realize their company is so slow that someone else is going to beat them to it, and decides to be that "someone".  Perhaps doing the design on their kitchen table, or over lunch breaks, they come up with a concept that they believe will create a sellable product, if not the "killer app" product.  At this point, he quits the company, often with a few friends, and they form a start-up company that gets to market first, wins a lot of market share, and completely skunks the former employer.  All of this by being free to make decisions, do what they think needs to be done, and usually by working 60 hours a week for a while. 

The big companies have those manuals of policies and procedures for a reason.  To borrow a saying, good decisions come from experience and experience comes from bad decisions.  The big company is so afraid of someone doing something wrong, they try to write procedures that will eliminate all chances of errors, which means they eliminate all independent thought from the day to day running of the company. 

Smart companies, however, are aware of this conflict between carrying out procedures and try to remain nimble enough to be competitive.  Fads such as the 1980's Japanese Management techniques (e.g., Quality Circles), and this decade's Lean Manufacturing (also out of Japan) sweep the industrial world regularly as managers strive to understand how to walk this balance. 

As another example, a large company might want to try and compete with a lower cost company that is trying to edge into their marketplace; for example, a Defense contractor making a GPS receiver might fear a commercial competitor's cost advantages.  They might want to change their entire approach to design of a product, taking more risk, instead of exceeding their requirements by large margins, they may try to just meet them.  Instead of a receiver that will work from deep winter in Antarctica to the worst of the Sahara, they decide to make one that would work for a day-hiker.  When they try to do it, the company's process manuals haunt them, and internal organizations, desperate to prove they are adding value, force the "low cost" program to adopt the same high-cost approaches they use in all of their products.  This destroys the attempt to lower the cost and ensures the failure of the program. 

Smart companies have addressed these problems by spinning off what are called "skunkworks".  The original skunk works (note it's two words, not one) was an offshoot of Lockheed, and their most proud achievement (at least, that they could speak about) was the SR-71 Blackbird.  A group of talented engineers went off in isolation from the rest of the company and created the fastest aircraft in the world, in record time, in the 1950s.  Skunkworks as a single word is the term used for such a group set aside by any company for a similar purpose.  This has been shown to be a successful technique, but note that the original company is still there, still working as an arthritic bureaucracy. 

SR-71 Production line at the Skunk Works

This is where we find ourselves as a nation.  

We are strangling in a bureaucracy with a Code of Federal Regulations that has grown like a bacterial culture.  A nation that was founded by a constitution that fills about 14 printed pages in today's technologies, passes financial reform bills that go over 2000 pages, health care bills that go almost 3000 pages, and more.   Each bill creates hundreds of new regulations, which are so poorly written they have to be refined by hundreds of court cases.  The court cases effectively create new law and new regulations.  Since congress is in session every year and passes at least one new law every year, the total number of laws and regulations increases without limit and everything eventually becomes illegal. 

What can we do?  We can't form a "skunkworks country" that can get around our laws and create a more mobile, productive society.  We only have one option: we have to create a national process, like industries do, to become more "lean, mean and low to the ground".  Get rid of superfluous laws.  We simply must reduce the size of the CFR and reduce the destruction caused by the regulation and litigation in our society.  To me, Tort Reform is absolutely essential.  A big part of the industrial lean activities is to study what policies need to be gotten rid of because "we've always done it that way".  The same should be done with the CFR. 

To be honest, there's more than one option.  The second option will result in blood in the streets, the deaths of thousands or even millions, and suffering on a colossal scale.  When companies don't handle their tendencies to be become too arthritic they go out of business.  The equivalent for the country is collapse, civil war and chaos.  Something to be avoided if at all possible.  Without reform, and without throwing out large chunks of regulations,  we are headed there.





Friday, February 21, 2020

How the Internet of Things Is a Throwback to the Middle Ages

It's an interesting perspective that I hadn't really thought of in those terms.  Now that I've read a law professor's opinions on this, I don't think it's far off the mark.  It certainly raises things to think about. 

Let me step back and set the story up a little more.  My day-to-day browser in the Windows world is Firefox.  One of the things Firefox set itself up to do is to offer me stories to read when I open blank tabs; those stories are from a site called Pocket.  Pocket appears to be an aggregator, because virtually everything it offers me is on site other than their own, but they actually do offer some content of their own.  In this case, “The ‘Internet of Things’ Is Sending Us Back to the Middle Ages” by author Joshua A.T. Fairfield, a Professor of Law at Washington and Lee University, is that content.

Fairfield opens by telling a story you'll probably recall; about a Las Vegas casino that was hacked through its fish tank.
The tank had internet-connected sensors measuring its temperature and cleanliness. The hackers got into the fish tank’s sensors and then to the computer used to control them, and from there to other parts of the casino’s network. The intruders were able to copy 10 gigabytes of data to somewhere in Finland.
He goes on to argue that the problem with Internet of Things (IoT) appliances like this fish tank is that we don't control them and it's not always clear who does. 
In my book, “Owned: Property, Privacy and the New Digital Serfdom,” I discuss what it means that our environment is seeded with more sensors than ever before. Our fish tanks, smart televisions, internet-enabled home thermostats, Fitbits and smartphones constantly gather information about us and our environment. That information is valuable not just for us but for people who want to sell us things. They ensure that internet-enabled devices are programmed to be quite eager to share information.
He goes on to talk about the Roomba issue that came up a few years ago - Borepatch covered it and while it was talked about, I don't think it was deadly bad P/R for the manufacturer, iRobot.  If you forget, the higher end Roombas were mapping owners' homes so that they could optimize their routes.  Early Roombas went in a line until they hit something, turned and went off in a more or less random direction.  The room-mapping units were more efficient.  Then it was leaked that iRobot was looking to share those maps of the layouts of people’s private homes with its commercial partners.

As Borepatch frequently adds, security wasn't an afterthought, it was never thought of.  Why should iRobot have any access to a map of your house?  Why should they be allowed to sell that information? 

Fairfield goes on to conclude a real problem is ownership.  You know that when you buy software, you're not taking ownership of it, you're licensing the ability to run one copy of it.  When you buy the latest snazzy smartphone, you're buying a useless piece of hardware because without the software it's running, it's not even a plain old phone: it's useless.  (We used to joke you bought a paperweight, but they're too light to be a paperweight!)

This idea is spreading to mechanical, physical things, not just software.
This sort of arrangement is destroying the concept of basic property ownership. John Deere has already told farmers that they don’t really own their tractors but just license the software – so they can’t fix their own farm equipment or even take it to an independent repair shop.
Divemedic over at Confessions of a Street Pharmacist has posted stories about Tesla motors essentially doing the same things, and turning off features in the cars' software if the car was bought used from a previous owner and running software not installed by Tesla themselves.  Nearly 15 years ago, there was a new name in home Digital Video Recorders: TiVo.  TiVo originally sold their DVRs by hard drive size and computer geeks quickly realized they could buy the cheapest DVR and replace the hard drive with one from their local store.  TiVo did their best to stop that before failing.

Where does the concept of the middle ages come in?  Back to Fairfield:
The issue of who gets to control property has a long history. In the feudal system of medieval Europe, the king owned almost everything, and everyone else’s property rights depended on their relationship with the king. Peasants lived on land granted by the king to a local lord, and workers didn’t always even own the tools they used for farming or other trades like carpentry and blacksmithing.

Over the centuries, Western economies and legal systems evolved into our modern commercial arrangement: People and private companies often buy and sell items themselves and own land, tools and other objects outright. Apart from a few basic government rules like environmental protection and public health, ownership comes with no trailing strings attached.

This system means that a car company can’t stop me from painting my car a shocking shade of pink or from getting the oil changed at whatever repair shop I choose. I can even try to modify or fix my car myself. The same is true for my television, my farm equipment and my refrigerator.

Yet the expansion of the internet of things seems to be bringing us back to something like that old feudal model, where people didn’t own the items they used every day. In this 21st-century version, companies are using intellectual property law – intended to protect ideas – to control physical objects consumers think they own.
If we truly own our own technology - I'm thinking our tablets or phones - we should be able to write software for them, or modify the Operating System if we feel we're qualified to do so.  The Android universe might be more accepting of that, but I know the Apple iOS people would do their best to stop anything that isn't in their App Store model.  Android for phones was rumored to be written by Google so that they can get their snooping technology into as many people's lives as possible - so they can harvest more data about users and sell it. 

Final words to Fairfield:
We need the right to fix our own property. We need the right to kick invasive advertisers out of our devices. We need the ability to shut down the information back-channels to advertisers, not merely because we don’t love being spied on, but because those back doors are security risks, as the stories of Superfish and the hacked fish tank show. If we don’t have the right to control our own property, we don’t really own it. We are just digital peasants, using the things that we have bought and paid for at the whim of our digital lord.


The iFixit Manifesto.


Thursday, February 20, 2020

10 Years of Babble

This post marks my 10th year of the blog!


If you want to be technical about it my very first real post was on February 21, 2010, tomorrow's date, but that's when most readers will find this post so it works out.  I've tended to let the timing on this post drift around a few days over the years.  I was playing around with the new colors and format, then realized I couldn't save it for tomorrow; it was either update it or do it over tomorrow.

As always, I thank you for stopping by.  According to the stats engine on Blogger, my number of views hasn't changed much in years; I get about 15 or 1600 per day.  The way those stats are provided makes it pretty much impossible to tell what people come here to read, or if they end up here by accident. 

As I said in the very first post, “I have no idea how long I'll be here, but here's to the ride!”


Wednesday, February 19, 2020

SpaceX Eying Water Tower Assembly for Better Ways to Build Stainless Starship

SpaceX is nothing if not a disruptive force in the aerospace industry.  From the hi-tech company model of prototype as soon as possible, fail early and correct quickly to the industry leading recovery and reuse of boosters, they seem to be always innovating and always pushing forward.  To borrow the cliche', they're always thinking outside the box.

I admit to missing this story on the 6th when Teslarati published it, but in the effort to improve the manufacturing of their stainless steel Starship, they're looking into the machines that build water towers to find ways to make the assembly faster and higher quality.  The first revision of this article included the following text:
A SpaceX engineer says that the company wants to adopt commercially-available manufacturing equipment that could allow its Boca Chica, Texas team to build Starship tank parts in minutes and nearly-complete rocket bodies in a matter of days.

Originally created to meet the needs of a variety of different companies – typically oil and gas related – that need efficient, affordable, and standardized storage tanks, a small but growing niche exists for semi-automated tank production. While there is some clear uncertainty given that the quality and consistency required for oil and gas needs or even simple water storage likely isn’t the same needed to meet strict spaceflight margins, SpaceX has already acquired several production tools from existing contractors and is working around the clock to prove that those same tools can be used to build large, reusable rockets.

The gamble is simple: if it turns out that off-the-shelf (COTS) equipment can become an almost turnkey solution for manufacturing high-quality Starship spacecraft and Super Heavy boosters, SpaceX may have found a shortcut to orbit, avoiding the huge expense of finding and building its own custom production solutions. But is that COTS tank fabrication hardware truly up to the task?
In response to that final question, Elon Musk replied to the Teslarati team with a lot of detail.  It now leads the article off.
This isn’t quite correct. An orbital rocket needs precision that’s 3X to 4X better than a water tower, so super precise parts, fixtures & welding are needed. Suborbital is much more forgiving.

That said, although substantial capital & engineering is required to achieve extreme precision, marginal production cost of the primary structure should actually be *less* than a water tower, because it’s built inside a factory in volume.

Unmodified water tower machines do not work well for orbital rockets, as mass efficiency is critical for the latter, but not the former. Hopper, for example, was made of 12.5mm steel vs 4mm for SN1 orbital design. Optimized skins will be [less than] 2mm [thick] in places across a 9000mm diameter. [Note: everything in square brackets my edits -SiG]
An interesting problem is how to support these 9 meter diameter steel skins that are 2 mm thick.  That's close to .080" thick across 29-1/2 feet diameter.  It ought to behave like an overgrown piece of paper with those dimensions.

The fact that SpaceX engineers are studying how the oil and gas industry gets tanks fabricated and how to modify those welding machines for the higher precision they need for spaceflight is interesting enough in itself.  If a solution to a similar problem already exists, it's generally cheaper and easier to adapt the old tool to the new problem than to invent an entirely new tool.


SpaceX's second test tank made it to about 50% higher pressure than it was required to survive, demonstrating that it's safe enough for manned space flight.  This was all hand-welded by very skilled welders on fixtures SpaceX developed.


Tuesday, February 18, 2020

Michael Flynn's New Attorney: FBI Actions 'Worse Than Entrapment’

Attorney Sidney Powell has taken on the case of Lt. General Michael Flynn, who served as President Trump's first National Security Advisor and was - by all accounts - railroaded into a guilty plea for lying to the FBI even though none of the agents who had spoken with him thought he was lying.  You might have seen her on TV if you watch lots of news; I recognized her instantly although I didn't really know her name or who she was.

Michael Flynn's attorney, Sydney Powell

She has come out with metaphorical guns blasting, according to this article from Free Pressers.  The first salvo she fired was in a private letter to Attorney General William Barr.
In the first paragraph of a private letter she sent to Barr, “Powell requested an outside auditor not associated in any way with former special counsel Robert Mueller’s team,” Rowan Scarborough wrote in a Feb. 16 report for The Washington Times.

“She argued that Mueller’s lawyers threatened Flynn’s son with prosecution to persuade the retired officer to plead guilty to lying to the FBI.”Powell wrote in the letter to Barr: “We request the appointment of new government counsel with no connection to the Special Counsel team of attorneys or agents to conduct review of the entire Flynn case for Brady material that has not been produced and prosecutorial misconduct writ large.”
“Brady material” is exculpatory evidence that both sides in a trial are obligated to provide for each other.

The letter apparently had the desired effect, as AG Barr has opened an independent investigation of the prosecution of General Flynn.  Mrs. Powell pointed out many facts about the investigation of General Flynn that's a collection of “things that make you go hmmm.”  Even casual consumers of news will recognize enough that just ain't right to question everything.
For one, she said, the FBI officially put Flynn under investigation in August 2016 in its Russian election interference probe but didn’t tell him during two encounters over six months.

An agent visited the Trump presidential campaign under the guise of a defensive briefing on Russian interference. In fact the agent was there to surreptitiously investigate Flynn, according to a report by Justice Department Inspector General Michael E. Horowitz.

Six months after opening the probe into Flynn, two FBI agents, including Trump antagonist Peter Strzok, on Jan. 24, 2017, visited Flynn at the White House under the guise of a friendly chat about national security. At that point, the FBI had no evidence of a Trump-Russia conspiracy outside of the Christopher Steele dossier, which Republicans say turned out to be a Kremlin hoax.

During the presidential transition, the U.S. had intercepted calls between Flynn and the Russian ambassador in which they talked about sanctions.
...
Inside the Obama Justice Department there was talk of investigating Flynn for violating the Logan Act, a 1799 law barring civilian interference in foreign policy for which no one has ever been convicted.
...
In her June letter to Mr. Barr, Ms. Powell recited her version of events. She said prosecutor Brandon Van Grack later acknowledged that the Flynn call was “perfectly legal.” It did not violate the Logan Act, which appeared to be the basis for the FBI visiting Flynn that day.

“There was no ‘Logan Act violation,’ and everyone knew it,” Ms. Powell wrote. “Yet General Flynn was illegally unmasked by the Obama administration and his call leaked to explode the ‘Russia collusion’ narrative in the press. The FBI interview was worse than ‘entrapment.’ He was led to believe he was having a casual conversation with friends about a training exercise from a day or two before, when in truth, it was a set-up-tantamount to a ‘frame, manipulated by Yates, Comey, Strzok, McCabe, and others to take General Flynn out of the administration. [Mueller team] then used it to pressure him to try to take out President Trump.”
It's a bit of a long and convoluted story, and I recommend you read the whole piece at the Washington Times because it goes into the parallel attacks against Carter Page and George Papadopoulos. 

Yesterday, Townhall Columnist Kurt Schlichter put up a piece called, “Burn Down the DOJ and Start Over” arguing it might well be too corrupt to save.  The only alternative is to burn it to the ground, salt the earth, and start over elsewhere. 
I want you to tell me, without bursting into laughter, that I am still supposed to respect our federal law enforcement institutions. I keep hearing about these wonderful keepers of norms and rules and stuff deserve our awe, and then I see the tawdry, self-serving and scummy way they operate, and gee – there’s a disconnect. A big one. If the price of our society is submitting to these corrupt and incompetent people of garbage, well, then I say burn it all down.
The more I read about things like this case, the more I agree with Kurt. 



Monday, February 17, 2020

SpaceX Launches 300th Starlink Satellite, But Muffs Booster Recovery

This morning, SpaceX successfully launched their fifth load of 60 Starlink satellites, reaching 300 satellites on orbit.  The only bad part of the mission was that the booster missed the recovery drone ship OCISLY (Of Course I Still Love You) and hit the water just off screen to the right of where the stationary cameras on the drone are aimed.


Screen capture from the SpaceX video coverage moments after the booster hit the water - the circled area is the cloud of water it threw up - a few seconds later, some of those droplets show up on the camera.  

While the mission is to put the 60 satellites in orbit, the booster (B1056) was a big part of the story for this mission.  First off, the booster last flew 62 days ago, making this a SpaceX record for the fastest turnaround to flight for a used booster.  A footnote to that record is that if they had successfully landed the booster it would have been their 50th successful booster recovery.  That milestone will have to wait.  Plus, that was B1056’s fourth launch in 10 months, which is unheard of with any other orbital-class rocket in existence.  The typical orbital class rocket still has a life of one use.

Since this was 1056's fourth launch, with a design lifetime of five launches, they're probably not going to try to drag it back into Port Canaveral from 390 miles offshore and spend the money/effort to refurbish it.  It would be worth knowing what the failure's cause was and especially if it was a result of deficient or defective work during that record 62 day turnaround. It's possible the reason is in data transmitted down during the flight and will be found by review of telemetry. 

Recovery of the fairings was another stated goal of the mission, but since I can't find mention of it anywhere I'll assume they were unable to catch either half with their two recovery ships. 

Another new thing in this mission was the launch profile.  Until today, the launch of these satellites followed a pretty standard template.  They launch and burn the second stage until they reach a spot in orbit where they turn the engines off and let the system coast upwards, gaining altitude for some time.  At the right time (I believe at the farthest point - or close to it), they fire the second stage again and circularize their orbit.  For this mission, they released the satellites into an elliptical orbit at a lower point for check out.  Once that's complete, small ion thrusters on the satellites will put them in their final circular orbits at 340 miles altitude.  This was said to require more out of the first stage booster and make recovery more difficult.

Eric Ralph at Teslarati has some good stuff here:
Based on the fact that B1056 kicked up visible sea spray just a few hundred feet from OCISLY’s deck, as well as the distinct lack of an obvious explosion, it looks likely that the Falcon 9 booster suffered some kind of navigational failure. It’s possible that it experienced the same hydraulic failure that disabled B1050’s four grid fins, but a new kind of failure – like anomalous GPS readings, a broken laser altimeter, failed Merlin 1D engine thrust vectoring, or something more complex – could be the ultimate source of the missed landing.

Regardless of whether parts or the entirety of the booster can be recovered, SpaceX will almost certainly learn a lesson (or several) from Falcon 9 B1056’s premature demise, hopefully allowing future rocket landings to avoid the same fate. Most importantly, today’s primary objective – placing 60 new Starlink satellites in orbit – was a flawless success, even if B1056’s loss is still a blow. SpaceX’s next Falcon 9 launch is currently scheduled no earlier than (NET) March 2nd and is unlikely to be delayed by today’s events.
My 2 cents worth: this mission was more visible than the last Starlink launch, but enough scattered clouds in just the wrong places to restrict it to maybe 20 seconds out of the 2 minutes or so we can see well.  Then I came in to watch the booster recovery and the near miss.



Sunday, February 16, 2020

Is Trump Peparing To Really Go After the Deep State?

Last week, the story got out that former White House Communications Director Hope Hicks would be returning to the White House, in a different role; to serve as a counselor to the president and senior adviser.  She will report to Jared Kushner. 
(Via Politico) […] White House press secretary Stephanie Grisham confirmed, calling Hicks “one of the most talented and savvy individuals I have come across.” Hicks departed the White House in March 2018 after working as communications director for Trump. She then moved to Los Angeles to work in a senior communications role at Fox Corporation.

[…] Hicks’ return to the White House gives Trump an ally who’s adept at translating his wishes to the broader staff.

Hicks was always well-liked among the communications and press staff, getting along well with the competing factions from the 2016 campaign and the Republican National Committee.
It's also being reported that John McEntee is returning to the West Wing after being fired in 2018 by then-White House Chief of Staff John Kelly.  McEntee is described as a “loyalist.”
In his post-acquittal initiative to finally clean the federal government of Obama holdovers and “never-Trumpers,” President Donald Trump has appointed a trusted loyalist to head up the Presidential Personnel Office, reports say.
...
McEntee is reportedly very close with White House senior policy adviser Stephen Miller from their days on Air Force One and the 2016 Trump campaign.

Administration officials told Axios that the president “feels he’s surrounded by snakes and wants to clear out all the disloyal people. Trump sees McEntee as the ultimate loyalist, and he has assigned him the powerful role of picking personnel across the federal government.”
The article linked above at Free Pressers gives their summary in the title, “No More Mr.Nice Guy.”  They believe Trump is going to surround himself with loyal people like Hope Hicks and John McEntee in an effort to start getting rid of Obama-holdovers and “Never-Trumpers” on the inside who have opposed him and slowed him down. 
Axios reported that “Trump has been asking for names of people he should fire. Many on the outside are more than happy to oblige.”
It's about damned time!

Red State offers this perspective:
Unlike a George Bush (either of them), Donald Trump arrived in Washington, D.C. with no connections to the DC policy establishment. In fact, he arrived after a bruising primary and general election that left him alienated from the GOP establishment. The GOP establishment might not have called his supporters “a basket of deplorables” but they nodded their heads and chuckled when they heard it. The same establishment had gotten rich and fat off illegal immigrant labor and outsourcing American jobs to wherever. They were used to keeping the GOP base in line with promises and crumbs (George W. Bush had GOP majorities for 6 of his 8 years, how much did they accomplish in regards to slowing illegal immigration or reducing abortion?) while delivering zero. When Trump arrived in Washington he was reliant upon the very same people who had opposed his election to staff his administration.  [Bold added: SiG]
Don't look now, but the left has tied the president in knots since he walked in - quite obviously planned from before the election.  Andrew McCabe was let off the hook last week.  Yeah, Trump was acquitted in his impeachment trial, but they're starting right back in and going after Attorney General Barr as well.  The entrenched deep State is winning.  Barr and Durham are two guys against thousands in the deep state.  What chance do you think there is it can be cleaned up? 

President Trump and Hope Hicks when she left the White House two years ago.


Saturday, February 15, 2020

Predictions About the Future Are Almost Always Wrong

But they can be kinda fun. One of the staple trade magazines in my old line of work is Microwaves & RF magazine (RF = radio frequency); one of their longtime editors is Lou Frenzel.  Lou has been in various trade magazines since 2005 and I've been aware of his work for most of that.  Lou writes a piece this month called “Perfect Vision - Predictions for 2020” and I thought his views on what's going on in the wireless 5G industry might be interesting.

I'll claim fair use and reproduce it here:
2020 clearly seems to be the year that critical mass is achieved, making it possible for the 5G wireless business to begin a growth spurt. Limited 5G service started in 2019. 5G has clearly won the all-time hype contest with continuous enthusiastic, over-the-top pandering, and untested declarations that may not be fulfilled. A rollout will take years, of course, but there has been a push by parts vendors, cellular operators, and consumers who want faster downloads of 8K movies and 24/7 cat videos to their iPhones. Here’s just a sampling of what to look for:
  • Fragmented and limited 5G cellular service in selected cities in the U.S. All major U.S. carriers will offer 5G services mostly in some of the major cities with a unique mix of both low-band (below 6 GHz) and high-band (above 6 GHz) services.
  • China will continue to dominate the 5G movement with an enormous number of users. The battle over the ban of China’s Huawei base-station equipment for security reasons isn’t completely settled in the U.S. and continues in Europe.
  • Initial smartphones will probably not have full coverage of both 4G and 5G bands, limiting their use mainly to local services. Some 5G phones may only cover the low bands and not the millimeter-wave (mmWave) bands. Others will cover both but will be far more expensive. Furthermore, you will need a new phone if you change carriers, as their operational bands aren’t the same. These phones will also continue to include the popular, more widely available LTE low bands. Connections over the mmWave bands may be spotty unless a user is close to one or more small cells that will make up the 5G infrastructure.
  • Fixed broadband 5G wireless service will be offered to consumers in competition with cable TV and DSL companies. Self-installed modems can produce up to 1-Gb/s data rates under ideal line-of-sight conditions.
  • ABI Research reports that subscriptions to video-streaming services are predicted to hit 91 million over the next five years. The speed of 5G will facilitate that trend. ABI also indicates that a looming major issue is the significant increase in energy required to provide 5G service. Who would have thought?  [Besides everyone - SiG]
  • The T-Mobile-Sprint merger, which appears to be getting the green light, could go either way. Combining both companies would produce a bigger enterprise that could more directly compete with AT&T and Verizon. However, if it does happen to still fall through, Sprint with take a downward spiral and eventually the assets will be acquired later, while T-Mobile will continue but remain a smaller third competitor.
  • Spectrum will remain a key issue in expanding 5G, but the FCC is working on the problem with auctions (C-band), spectrum swaps, and band sharing (3, 5 GHz). Initially 5G will mostly use spectrum below 6 GHz. A few mmWave systems will emerge. Some predict that carriers will phase out current 3G spectrum faster than the 2G shutdown to make way for more 5G bandwidth.
  • A growing consensus feels that 5G will be essential for two key technologies—the Internet of Things and self-driving cars. With the growing number of IoT products and installations, 5G will have the capacity to handle the massive amount of data. Autonomous vehicles are expected to need the much lower latency that is reportedly available with 5G.
Bruce Lancaster of Wilson Electronics indicates that as 5G rolls out in 2020, the current LTE network will stay in place to serve 5G until the more advanced network is available. He believes that its primary use will be the expected voice and data services. The “killer app” for 5G has not emerged. Customer capacity will grow. The low-band spectrum will dominate (e.g., 850 MHz from AT&T and 600 MHz from T-Mobile) initial systems. A major move to mmWave bands will come later. Mr. Lancaster feels that even as 5G comes on line, Wilson’s line of repeaters or range extenders will still be needed to ensure more reliable connections, especially in buildings.
There you have his take.  Is this absolutely likely to be how it turns out this year?  I doubt it's exactly right, but I wouldn't be shocked if he ends up 80% right.  Most of these problems are pretty well known, but the 5G hype is everywhere. 

My take is that 5G is going to be a big nothing for most people who buy the shiny, expensive new phones.  8K videos?  I'm going to go out on a limb and guess there's 100 8K videos in existence in the world, and most people will be interested in watching a couple.  If their cat videos or social media loads faster, the novelty will last an hour or two.  If the network is good, 4G LTE can do over the Gigabit/second barrier now, supposedly 5G speeds, but only newer phones with better hardware can do it.  4G LTE just doesn't have the buzz behind it. 

But, after all, industry is already laying the groundwork for The Next Big Thing, 6G, to roll out in about 10 years.


“5G??  Dude, those are power lines!”  Ah... yeah, but if you look closely just left of and below center, you can see what looks like a cell tower. 


Friday, February 14, 2020

Dropping a Perfectly Good Space Capsule Out of an Airplane

I'm sure everyone has heard that skydiving joke about jumping out of a perfectly good airplane. Teslarati reports that SpaceX is just about done with a round of tests of the parachute systems that will be put onto the Crew Dragon manned capsules, and could have just two tests left.  The tests began last October in response to several major failures that occurred during drop tests both in 2019 and 2018.  These failures led SpaceX and their contractor, Airborne Systems, to redesign the parachute system.  These sorts of tests are intended to test the parachute system at its worst case limits - better to fail being pushed out of a C-130 with no one in a box than with astronauts on board.  The tests appear likely to be completed before the end of March. 
Most recently, SpaceX has been aggressively testing the latest Mark 3 (Mk3) parachute variant with great success and has completed some two-dozen consecutively-successful drop tests since October 2019. Now, NASA and SpaceX are working together to settle on a design for two final Crew Dragon parachute tests, the results of which will almost certainly determine when the spacecraft’s astronaut launch debut will occur.
They go on to report that SpaceX has “quite literally been pushing the envelope of parachute engineering and the immensely complex physics behind their behavior during deployment.”
In response to the additional testing and analysis NASA required after a recent April 2019 chute failure, SpaceX has essentially been forced to push the state of the art of parachute design and modeling to new levels. NASA says that SpaceX has begun to model certain conditions and newfound failure modes in ways that “provide a better understanding of parachute reliability” and have forced NASA to reevaluate its own standards and certification processes. Shown in the video above, SpaceX recently completed a successful parachute test identical to the attempt that failed in April 2019, a major step towards confirming that the new parachute analysis and design have mitigated prior faults.” 
The Teslarati reporters say that SpaceX says they've completed no less than 24 successful Mk3 parachute tests, ranging from single-chute and chute-out tests to the full-fidelity spacecraft launch and recovery that followed Crew Dragon’s January 19th, 2020 In-Flight Abort (IFA) test.

While these tests complete over the next few weeks, Spaceflight Now reports that the Crew Dragon capsule slated to carry NASA astronauts Doug Hurley and Bob Behnken to the ISS arrived Thursday at SpaceX's facilities on Cape Canaveral.  The first manned SpaceX mission to the ISS is still slated for May 7. 
In a statement, NASA said the Crew Dragon will be the first spacecraft to launch astronauts from U.S. soil since 2011, when the space shuttle was retired.

“The SpaceX Crew Dragon spacecraft for the first crew launch from American soil since 2011, has arrived at the launch site,” NASA said in a statement. “NASA and SpaceX are preparing for the agency’s first flight test with astronauts to the International Space Station as part of the agency’s Commercial Crew Program.”

In 2014, NASA tapped Boeing and SpaceX with contracts valued at $4.2 billion and $2.6 billion, respectively, to develop, test and fly commercial human-rated spacecraft designed to ferry astronauts to and from the space station.
Hurley, a pilot on two space shuttle missions, will serve as vehicle commander on the Crew Dragon test flight, known as Demo-2. Behnken, also a veteran of two shuttle flights, will be the vehicle pilot.

If these last few test drops result in changes to the Mk3 parachute design, it's my bet the system can be fixed here in Florida rather than send the capsule back to Hawthorne (CA). 


 SpaceX photo.


Wednesday, February 12, 2020

Quarantined Cruise Ship in Japan May Now Be The Best Place to Study the Virus

The news broke today that the cruise ship Diamond Princess that has been in the news for being quarantined in Japan now has 175 confirmed infections from the 2019 novel coronavirus (2019-nCoV).  There are 39 new confirmed cases among passengers and crew members and one confirmed case in a Japanese quarantine worker.
Since the outbreak began in December, there have been over 45,000 confirmed cases of COVID-19 worldwide and at least 1,115 deaths. But while 2019-nCoV has spread to at least 24 countries beyond China, nearly all of the COVID-19 cases and all but one death have occurred in China.

According to the latest figures from the WHO, 44,730 cases are confirmed in China, while a remaining 444 are outside the country—including the 175 cases linked to the Diamond Princess. The tally is by far the largest outside of the outbreak’s epicenter; the country with the next-largest COVID-19 outbreak is Singapore, with 50 confirmed cases, according to WHO.
It's reported that the Japanese quarantine worker wasn't wearing highly protective clothing for high biosafety level areas, just wearing a mask and gloves.  He was handing out questionnaires and checking the health of passengers and crew members. 

Assuming passengers don't jump overboard and find ways to get back onshore (always easier at a dock than when miles at sea), this seems like a good opportunity to study the virus.  It could allow study of how long the virus survives on surfaces, and how effectively it spreads.  This is an isolated population, after all.  Patient zero, the first patient to contract the virus on the ship, is known:
Screening for COVID-19 on the ship began after a previous guest tested positive for the virus in Hong Kong on February 1. The man, who is from Hong Kong, boarded the Princess January 20 in Yokohama at the start of a 14-day round-trip voyage. The man sailed a leg of the voyage before disembarking during a stop in Hong Kong on January 25. Meanwhile, the ship sailed on. Upon news that the guest tested positive on February 1, the Diamond Princess returned to Yokohama a day early and has been quarantined ever since, with guests in isolation in their cabins.

It is still unclear when and where the man from Hong Kong became infected and how the virus has spread among people on the ship. It may be that the outbreak involved a so-called “super-spreader,” which means that a single infected patient sheds the virus extremely efficiently and infects an inordinate number of people. But again, it is unclear how many people may have brought 2019-nCoV aboard and how—or if—the 175 cases are all linked.  
I suspect that most of you have been reading what Aesop at Raconteur Report posts about the virus, like this one.  I sure have.  Having an isolated population like this might be a good way to study this virus and perhaps reduce the uncertainties that seem to apply to just about every characteristic.  There are 3,711 people on board the Diamond Princess; Reuters reports that roughly 80 percent of the passengers (over 2,900) are aged 60 or older, with 215 being in their 80s and nearly a dozen over 90. Those age groups have been among the most vulnerable demographics in the outbreak overall.  If those same percentages apply to the 175 cases and 80% of the cases (140 people) are over 60 years old, I'm afraid that will be where the largest number of deaths come from. 


The Diamond Princess at dock in Yokohama.  Getty Images photo.