Sunday, June 11, 2017

Arguing The Point

Someone on the internet is wrong. You wish to let that person know they are mistaken, as well as inform others to be sure they too do not mistakenly believe this to be true. How do you argue your point in an effective manner, disagreeing with someone, while maintaining civility? There's a good post on How to Disagree by Paul Graham which lays this out, and I'd like to touch on this, especially as it relates to science and engineering discussions.

We're living in a world where we are surrounded by incredibly complicated technology - sometimes the simpler that tech looks on the surface, the more complex it is. There's a large percentage of the population that would have a hard time explaining technology as old as the internal combustion engine in your car, and things get worse from there as you move through things we all use every day but barely know it - encryption and compression get the latest episode of 'House of Cards' onto your TV, but care to lay out how that happens for me? 

More importantly, how do you tell when someone is making false or exaggerated claims about technology? The facetious answer is "study for long enough to become an expert" however even then it's hard - there's a lot of technology out there, and only so many hours in the day. If you don't know the technology, and people are arguing about it, how do you evaluate the arguments they make? One way is to evaluate how they are arguing, even if you don't understand fully what they are saying.

The pyramid in the graphic above is an attempt to lay out Paul Graham's hierarchy (I have no idea where it's from, I'm not taking credit for it, thanks to whoever did it). The pyramid you see here contains the 'best' arguments at the top, the 'worst' at the bottom, and as you can see, it's wider at the base to represent that it's a lot easier to make the worst arguments than the best - 80% of everything, after all, is crap.

At the base it's fairly obvious - if one side is saying "the other guy is a doo-doo head" then they don't have much on their side. Sadly, you don't have to go to far above this for most people to lose critical thinking and an ability to evaluate what's being said. Ad Hominem is actually quite effective in discrediting a party with some audiences - for example saying "they're just a disgruntled former employee with an axe to grind" while ignoring any detailed points that person may have made, or whether they are even justified in being disgruntled. The most common form of argument is often simple Contradiction with no evidence to support it - "My client is innocent, and we're confident that the jury will agree." Frustratingly, we seem to be in a world where there's a media bias to he-said/she-said and placing the weight of argument 50/50, regardless of actual merit of the case.

I've never seen this fake-balance more brilliantly demonstrated than by John Oliver in this Daily Show segment on the Large Hadron Collider, and how there is a "50/50" chance the world would end when it was switched in. The part in question is at around 3:00, but I'd encourage you to watch the whole thing, it's John Oliver comic genius.

Dr Ellis is my hero here. Watch him start at the 'top of the pyramid' and refuse to be dragged down into what we're used to from media. He stays on topic, doesn't get tricked by Oliver (way harder than you might think), sticks to his point, and doesn't let himself be drawn into the arguments from the base of the pyramid. It's a short segment but really highlights how awful the media are in pushing junk science from those with limited understanding compared to those with deep knowledge - but where's the audience in that?

To further illustrate that point, and to show an example of an argument on a technical matter, let's take with this recent statement from uBeam 

uBeam is an innovation that will breed innovation. Ubiquitous wireless power will lead to a world with smaller batteries and thinner, lighter devices. With wires virtually eliminated, TVs can sit in the middle of a room cord-free and light fixtures will become “stick-on” without the need for routed power. uBeam is also a universal standard, making those bulky travel adapters a thing of the past. Imagine charging your phone, laptop or even your hearing aid virtually anywhere, without any effort. This is life powered by uBeam.

I'll take that one bolded point - that TVs can be powered wirelessly with ultrasound in the middle of a room and try and 'refute the central point'.

I'll begin by trying an argument against it:

"They're a bunch of stupid poopy heads" - No, that's bad, that would be Name-Calling

"They're just disgruntled current employees desperate to share their misery with the prospective employee and have no idea of the basics of physics" - No, that's Ad Hominem

"Powering a TV with ultrasound in the middle of a room is not a practical possibility, and is around one hundred times larger a problem than charging a phone in the same manner. While it is theoretically possible, the costs, inefficiency, and safety concerns are staggeringly high, while practical alternatives are low cost, and there is no economic demand to make this happen. Regulatory limits make it difficult in the US, and impossible outside the US." - OK, now we're doing somewhere between Contradiction and Counter Argument.

Let's move this to an argument from the 'top of the pyramid' by Refuting the Central Point

I'll begin by stating my assumptions:

We're talking about a large screen TV, not a small hand held. The TV is in a room you have some control over the infrastructure. The TV does not have a battery and needs a constant supply of power to work that can't be interrupted. From Energy Use Calculator I'm going to take 100 Watts as the power requirement for a 50 inch LED TV. We'll be assuming this is in the USA, and that the pre-2015 OSHA safety regulations are in effect and that in no location is sound over 145 dB used. Outside the USA the 115 dB limit give a transmitter and receiver 1000x area increase requirement. Note this will also apply within the USA should current OSHA limits restrict usage to 115 dB.

I will assume a generous 33% efficiency on receive, with 50% efficiency from transmitter to receiver incorporating both distance and angle of incidence. I will assume there is infinite power available into the transmitter and that efficiency from the wall socket to the ultrasound conversion is also 50%.

I assume each phone case sized receiver, at 5 by 10cm, uses $10 in parts, and we need to sell at 3x BOM to make money.

I'll ignore nonlinearity for the sake of simplicity, even though that's likely to become an issue, and limit the separation of transmitter and receiver to no more than a meter.

Now my calculations:

100 Watts powered means 300 Watts acoustic needs to be received. (100 Watts at 33% efficiency). This compares to around 0.5 Watts requirement for a phone, hence the "hundred times larger" comment. At 145 dB ultrasound is around 300 Watts/m2, meaning the receiver will need to be 1 m2 in size, that is a square of 1 meter on each side, or equivalent. A 50 inch TV is around 25 by 44 inches in size, (64 by 112cm or 0.72 m2)  so as a meter is around 40 inches, that means the receiver will be around 1.5 times the size of the TV. Ooops, better get to work on that efficiency.

Now a panel that's 1m2 is about the size of 200 phone cases, so around $2000 in parts, or $6000 in cost to sell and attach to that 50 inch TV, that costs around $500 right now.

The Transmitter needs to be twice the size of the receiver to take into account that 50% efficiency, so it's 2m2, and from the above calculation that means $12,000 for the transmitter.

Note that if the regulatory limit is 115 dB then the area scales by a factor of 1000 and the transmitter and receiver are each larger than the room.

For power supply, going with the efficiencies, the wall socket needs to provide 100 Watts, times 3 for the receiver efficiency, times 2 for the transmit efficiency, times 2 for the conversion efficiency, for a total of 1200 Watts. Fortunately this is (just) what a 110 volt 15 amp circuit can provide at 80% max load regulations allow (1300 Watts).

At 5 hours usage per day, and 12 cents/kWh average power cost in the USA, that's 72 cents per day to run, or $262, of which $22 is the actual TV use, the rest the wireless power system.

The additional 1100 Watts to use the wireless power system will be lost as heat (it's about a one bar electric fire equivalent), so in the winter that will save money, in the summer you'll need AC. I'll call it a wash to simplify this.

Summarizing the Argument:

Given the above, to power a TV wirelessly with ultrasound, it will cost $18,000 in the transmitter and receiver, with an additional $240 per year in running costs. Assuming efficiencies are as high as stated. And that no-one walks into the beam, since any interruption will make the TV switch off. And that you don't mind a receiver that's larger than the TV. And a transmitter that's twice that size and isn't too far from the TV. And that the room gets a bit warm. And that it's in the USA and the OSHA limits don't change to match the rest of the world.

But other than all that, isn't that much more awesome than running a $5 cord to the nearest outlet or paying someone to run a cable under your floorboards?

I think I'm going to call this "impractical".

OK, sarcasm over - I've run my calculations, providing all assumptions, workings, references etc so that anyone who disagrees can say "Your assumptions are faulty, here's what they should be" and then it's simple job to rerun those calculations get the new numbers, and judge from there. If anyone who is an advocate of wireless power would like to argue with these, feel free to correct me, and let's see where it takes the numbers. Or argue that my methodology is incorrect, I'm happy to do so - but like every other time in this blog where I have presented numbers, equations, and physics as core to my argument, I expect I'll be met with silence or more questions as to my motives. The top of the pyramid meeting with a response from the bottom.

My point to most people is this - if you don't understand the physics or details of a technical discussion like this, look to those presenting actual data, references, and their methodology and assumptions. If there is one side doing that, and the other calling names and questioning character, then you should likely consider one side's argument as superior to the other. If both are arguing methodology and data, then you may be watching a genuine scientific debate, which is good and healthy, it's what we want. If both are calling names, they're both idiots.


  1. There's a 100% chance you're my favorite blogger right now.

  2. This comment has been removed by a blog administrator.

  3. This comment has been removed by a blog administrator.

  4. Nice work.
    Your calculations are of course overly generous, but that's the interesting part. If you can debunk something as being completely impractical using overly generous (non practical) best base numbers as you have done, then the actual practical numbers are going to be so horrifying were no amount of extra effort, more funding, or tweaking the design will make any difference at all to it being ever practical.
    Nice work.

  5. i really likes your blog and You have shared the whole concept really well. and Very beautifully soulful read! thanks for sharing.
    GCLUB มือถือ