omenti

0 → 1 → 2 → 3? → 4 → → → → 5

A curious Bell System business from the late twenties through the thirties was selling bundled outbound courtesy calls to the airlines. The public saw flying as risky and it was common for an airline ticket to include a free one minute phone call let your family know that you had made it to your destination in one piece. It wasn't great pr, but it was an appreciated courtesy. 

 

Until the fifties airlines had a serious safety image problem.  Flying had been unsafe.  Going up in one of those caniplicons was seen as taking your life into your own hands. Fine as long as you didn't kill anyone else..  The same was true for the airplane. No safety standards existed for pilots or airplanes.  The emerging airlines stayed afloat largely with airmail contracts with the post office.  They pushed for safety and training regulations and tried to create an image of military-like professionalism with uniforms and codes of conduct.  At the same time enormous progress was made in airframe, engine, instrument and control system design.  Aerodynamics became a serious area of study during World War I with the creation of NASA's forerunner - The National Advisory Committee for Aeronautics.  From the biplane to pre-jet fighters and bombers in twenty years.  

 

Flying was doubly dangerous as many of the aircraft of the day weren't stable and pilots weren't very skilled.  The autopilot emerged and was soon lionized as making flight safe and easy.¹  In fact the technology slowly emerged along with everything else.  I'll skip the detailed history and focus on something just as interesting that is relevant to automated systems today - the reality and perception of accountability in complex human-machine interfaces.

 

Autopilots grew in sophistication though the thirties with a focus on flying an airplane when visual flying (VFR) wasn't possible.  After WWII serious experiments of hands-off flying were made including a 1947 flight from Newfoundland to England using a system that read a series of punched cards.  

The plane behaved as if an invisible crew were working her controls. … The commanding robot was a snarl of electronic equipment affectionately known as "the Brain." Everything it did on the long flight was "preset" before the start. It received radio signals from a U.S. Coast Guard cutter. Later it picked up a beam from Droitwich, England, and followed that for a while. When the plane neared Brize Norton, the wide-awake Brain concentrated on a special landing beam from an R.A.F. radio and made a conventional automatic landing. On the way over, the crew checked the course and watched the instruments. Most of them had little to do. They played cards and read books.

Time 1947

 

The technical press were beside themselves.  They saw a jump to a future where the pilot was completely out of the loop.  The emergence of remote controlled military drones muddied the water a bit as some reported them as autonomous.  There were predictions that pilots may be an endangered breed and the very unstable helicopter would be rendered easy to operate with full or near autonomy and would quickly spread to every middle class garage. 

 

It turned out to be a tad optimistic, but terrific progress continued with the average commercial flight only needing pilot intervention on takeoffs, landings and while taxiing.. and even then takeoffs and landings are easily automated.  

 

The systems grew in function and complexity. The software effort has grown to the point where the software component of a new fighter or airliner often exceeds half the total development cost. Designing for and testing and integrating complexity is a non-trivial task.²

 

There is the issue of what happens when something goes wrong.  Almost always some human in the system is blamed.  There may be a series of events that could have seen pilot intervention, but didn't - or perhaps saw the wrong intervention.  Pilot error...  The systems and their designers are rarely blamed in aviation.  In the early days autopilot and instrument makers would settle lawsuits out of court rather than take a reputation hit.   Detailed analyses of crashes often center around the crew not being able to spool up quickly enough to realize what is going and and how to do something about it.  More often than not the automatic system encountered something outside its design envelope (sometimes they fail within their envelope) and the passing of command fails. 

 

It turns out these events aren't that uncommon, but the crews have enough time to react.  The fundamental problem is our conscious minds mostly single task and are interrupt driven.  If I'm not directly flying the plane and the computer says - here Steve, it's all yours - I need fifteen seconds to a minute to come up to speed.  When something awful happens we blame the crew.  It is as if they are human airbags that deploy to protect the technology.

 

And that brings us to self-driving cars.

 

Autonomous cars are a much more difficult problem than autonomous aircraft.  About a decade ago serious progress was made that since has been fueled by the addition of cubic money with many expecting big things in the short term (say five to ten years).   To sort things a six level automatic taxonomy from the Society of Automotive Engineers is useful

 

level 0  no automation.  A car from the 70s and before (some would argue automatic transmissions are level 0.5)

level 1 single task driver assistance.  Cruise control, stability control, anti-lock brakes.  Universal on most current American cars.

level 2 partial automation.  Two more more subsystems combine.  The combination of cruise control and lane keeping for example.  Common on many cars today.

level 3 conditional automation. Systems that drive and monitor, but assume human backup.  Tesla is sort of an example

level 4 high automation. Systems that do nearly all driver tasks, but only in certain regions and conditions.  Interstate highways in daylight and dry weather, or a very low speed urban system that is tailored to just that environment.

level 5 Full automation. The works.  It would go anywhere you can imagine going on or off road and in all conditions. No driver controls are necessary.

 

Currently we're at level 2.  Tesla assumes a human in command with their level 3 system, but the human attention recapture issue rears its head.  Many, myself included, believe it is sufficiently difficult that level 4 happens before we have solid level 3 systems.³  Level 4 in certain carefully defined environments seems likely in five years.  I've done a bit of work with some folks looking at the system and human level issues of such designs and note it gets thorny in a hurry.  Some may plow ahead and use human airbags to protect their growing technology base, but that may be counterproductive for the industry.

 

Widespread level 4 will be a long rollout like autopilots in aircraft were.  The software is enormously complex and some of us believe some breakthroughs in software verification and testing are needed in addition to a much deeper than just the technology view of the problem.  Expect certification and regular maintenance requirements.  It may be that personal ownership won't be practical if maintenance requirements mirror the aircraft world.  This isn't just rocket science - that's the easy part.  Level 5 is way out there even if the tech exists.  We could have it on airplanes but won't go there for social reasons.  

 

In the end I'm optimistic on highly constrained level 4 systems, but I'm definitely not in the  technology solutionist camp.  If Apple, Toyota or any of the other players are doing this I can see it in cooperation with existing transportation systems - probably in Europe or China - and not privately owned.  I'm also very optimistic about potential lower death counts, but much of that is the kinetic energy argument as speeds are likely to be slower.  There is a lot of detail here, but beyond the scope of an hour blog post.  So many interesting scenarios...

__________

¹ The article is from the New York Times, 27 Nov, 1916

² We complain that air traffic control could be much better. It could, but the dose of reality is that changes have to propagate across many systems and they need to be checked. We could rely completely on GPS, but what happens if there if the system goes down (not terribly likely, but possible) or if there is local GPS denial (this has happened)? Dozens of systems need to be in synch and the mentality of careful testing and verification that the airlines desired beginning ninety years ago continues.  No one wants to tamper with the safety record story.  I don't blame them.

³ Google seems to have taken this point of view.  Their focus is now on highly constrained level 4 systems.  They are very worried about the human attention recapture problem.  The same is true for most of the Detroit and Euro systems I'm aware of.

 

__________

 

 

congratulations graduates and welcome to the battle

One of my most significant and hard-won learnings over the past ten years has been that conventional education is useless - even counterproductive - in combating pseudoscience in a polarized society.  I've touched on it hear and a few of you have heard far too much from me.  So rather than more from me here's Atul Gawande's commencement address at Caltech last Friday.  For those who prefer print (I usually feel something is lost), the transcript appears in The New Yorker ..

Caltech commencement addresses are often good.

batteries

A somewhat past prime lemon was sitting around making me wonder..  Let's say you're a Florida citrus baron and your car's battery is run down.  You remembered the lemon battery you helped your kid make for a sixth grade science project and wonder how many lemons would it take to crank the car?

 

A lemon battery is pure simplicity.  Push a galvanized nail and a copper wire into the unsuspecting piece of fruit and connect it to something like a light emitting diode current flows through the completed circuit and the bulb glows.  The flow of electrons from the zinc to the copper only happens when the circuit completes.  In theory the battery could provide power as long as there is zinc and an acidic enough environment to keep the reaction going.  In reality there are many factors that shut things down before the zinc is completely dissolved.   

 

I had never made the measurement.   A minute playing around and the Zn-Cu battery measured 0.82 volts at 1.2 milliamps - just under a milliwatt.   A average car starter needs about 3 kilowatts of power to crank an engine.  Assuming you have some way to connect your fruit without loss, you'll need about three million lemons.  Calling AAA is a bit more practical.   Lemon batteries score high as science education demos, but low on almost every other battery metric you can imagine.

 

You can almost go to the periodic table, pick a couple of elements, and make a battery. Most will be even worse than even our lemon Zn-Cu battery, but a few will stand out in one way or another.   Depending on the application you are interested in six major criteria: how much energy can be stored (energy density), how much power can be delivered (power density), lifetime (how many charge/discharge cycles), charging rate, cost, and safety.  Unfortunately no single chemistry is optimal across more than one or two criteria.  But then good engineering has always been about finding a good-enough optimization.

 

 

Smartphones and cars make us think of lithium ion batteries.   If you're building a space probe nickel-hydrogen is your friend.  It doesn't match lithium-ion in most criteria -  it costs several hundred times as much per kilowatt-hour - but it can go through tens of thousands of charge-discharge cycles.  Not having to replace a battery after a few hundred cycles when you're in deep space makes it seem cheap.  Flow batteries are relatively inexpensive and can go through a large number of charge/discharge cycles, but have a low energy density that relegates them to stationary use. Some lithium-ion chemistries have low costs, but tend to be highly flammable. Power densities - how quickly energy can be transferred from the battery - vary widely ... a very important automotive consideration. Energy and power densities are stronger functions of temperature in some chemistries. This may be ok if your battery is in an indoor power farm, but a serious problem in cars and smartphones.  Dozens of other battery chemistries have their niches, but a lithium-ion comes closer to satisfying more of the criteria than most competitors for many consumer applications.

 

Lithium-ion is a catchall term describing dozens of chemistries. It only tells you the charge is carried by a lithium ion.  The important bits that define a battery - its anode, cathode and electrolyte - are not specified.  The graphite-cobalt oxide battery in your smartphone has very different characteristics from the graphite-lithium manganese oxide battery in your portable drill.  The battery in your smartphone is not ideal for automotive use, but the fact billions are manufactured made them cheap enough to balance some of their other problems.  Tesla felt the wide availability of consumer electric class batteries was enough of an advantage that they could accept some of its issues.  So far that seems to be working well.

 

 

Although smartphones are the current major user of lithium-ion batteries we'll see a transition to automotive use rather soon.  A plug-in EV requires about 5,000 times as much power storage as a smartphone.  Two hundred thousand EVs is about a gigaphone of batteries.  For automobiles it is far from clear which chemistry will win in the short or medium term.  The hope is that entirely different chemistries will become practical in the long term.  I'm also hoping for improvements in smartphone class batteries.  We may see a factor of two in power density in the next decade, but I'm getting ahead of myself.

 

In principle a lithium ion rechargeable battery is simple.²  Generically they consist of two electrodes separated by a thin membrane and an electrolyte that allows ions to freely pass from electrode to electrode.  When the battery is charges lithium ions are released from the positive electrode, the cathode, which is a lithium alloy such as lithium cobalt oxide.  The ions move towards the negatively charged electrode, the anode.  Anodes are usually made from graphite and the ions find spots between the carbon atoms to settle in.  This process requires power electric power to drive it.  

 

To use the power just complete the circuit  by connecting the electrodes to something like a motor.  The lithium atoms embedded in the graphite give up electrons which travel through the circuit running car's electric motor.  At the same time lithium ions leave the graphite and travel through the electrolyte and membrane to the cathode where they meet up with electrons that have made the trip.  That's it - a nice reversible system.  

 

Graphite is used as an anode largely because its a good conductor, but it isn't efficient at storing lithium ions during charging - it takes six carbon atoms to hold one lithium ion.  Silicon seems more attractive - in theory one silicon atom can hold four lithium ions.  A silicon anode can, in theory, store much more energy.  Unfortunately material science hasn't come up with a good way to make a stable anode that can take many charges and the rigors of movement.   Well - perhaps that is temporary.  There have been a few exciting laboratory developments that might yield at least a factor of two better energy density and perhaps at a lower cost.  You could double the range of the Bolt or Model 3 to four hundred miles. Assuming one of the developments work out a viable automobile battery is probably a decade out, but there are signs of progress.³  Currently lithium-ion batteries have about 1/80th the energy density of gasoline.  In automotive use it isn't quite as bad as it sounds as electric drive trains are more than three times as efficient as internal combustion drive trains and you can get away with some extra weight.  A factor of two would be an enormous improvement that would answer most range anxiety worries.  Do it at a low enough cost and you have to worry about ramping manufacturing and charging infrastructure quickly enough to meet consumer demand.

 

The long term may move beyond lithium-ion chemistries entirely.  Metal air batteries use metals like aluminum, lithium, magnesium, silicon and the oxygen in air along with aqueous electrolyte.  In theory they have energy densities approaching gasoline when you factor in the efficiency of the electric drive train. Non-rechargeable primary batteries have been built, but a practical rechargeable is only a dream at this point.

 

__________

¹ If you've done it right citric acid causes  zinc on the nail into the lemony pulp as charged ions leaving electrons behind in the metal.  ( Zn → Zn²⁺ + 2e^- )  Positively charged hydrogen ions in the pulp, technically the electrolyte, combine with a couple of electrons at the copper surface forming hydrogen ( 2H⁺ + 2E^- → H₂ ).  The oxidized state of Zn has lower energy and the energy released provides the power that flows through the circuit lighting the light emitting diode.

² Rechargeable batteries are called secondary batteries, non-recharcables are primaries.  Although primaries often have much better energy densities than secondaries, replacing them ranges from very expensive to impractical.  An iPhone 6 Plus battery stores about 15 watt-hours - that would be nine AAA alkaline batteries.  The iPhone's battery should be good for about three years of charges.  If you need to recharge your phone every day you'd pay about eight dollars a day for three years - $2,800 - just for batteries! Imagine the landfill and production issues.

³ I'm partial to a novel silicon approach that uses nanoparticles.  It should be stressed there are several very serious approaches underway with a dizzying press release schedule.  Most of them will not amount to anything commercial, but are adding to the knowledge base - at least those that are published.   There is no guarantee solutions exist.   An area where software is not eating the world...

 __________

Recipe Corner

 

Nothing to report foodwise other than my seasonal shift to my pressure cooker.  They're safe and much faster than normal cooking - just the trick for keeping the kitchen cool.  

The reason they work is not their higher temperature (250°F is not that much above boiling).  Steaming is efficient as a huge amount of energy is released in the  form of latest heat.  But the mechanism is often confused ...  It isn't forcing liquid into the food - that turns out to be a very slow process.  It turns out to be the density of the steam.  Density increases by a factor of two so the steam contains twice as many water molecules to deposit their latent heat.

You don't need to know how a pressure cooker works - just know modern cookers do a fabulous job.  I even have hiked with a lightweight model.  When you're in an area where you have to pack your fuel, being able to speed up cooking dramatically quickly reduces your fuel need and the total weight of your pack.