More uBeam at CES 2019 - MacVoicesTV Interview

There was a brief video published today interviewing the uBeam Director of Product Marketing and Management at the floor show in CES (not the private suite), by MacVoicesTV. What you see is quite limited, a receiver only, no transmitter (except off camera, and on the screen behind), but interesting nonetheless.

The Interviewer

There are a few things that stand out in this video, and I'll start with the interviewer. Now I know this is not meant to be an in-depth technical expose, I'm eternally frustrated at the terrible coverage tech gets, but this was little more than allowing uBeam to read out their marketing material unchallenged. What is the point of the interviewer if they don't add anything to the mix? He does say "I wouldn't be doing my job if I didn't ask about safety" and then lets them go wild with, IMO, very suspect safety statements (more on that later). After that, nothing that his job actually entails such as asking about power transmitted, delivered, cost, efficiency, product release date, regulatory approval - anything of practical interest. He was basically pointless, and should have just handed the mike to the PR guy and cut the pretense he was doing anything at all.

The Demo

Now the demo itself did yield some interesting info. In the background the poster claims "Useable Power, Meters (away?)", so definitively claiming >1 meter here. The receiver shown lit up - an LED - so a whole few mW of power has been received. Distance from the transmitter is not shown, but likely a meter or so. So much for phone charging if that's all they can do, and you have to think that if they could do more, they absolutely would be showing that.

For tracking, the target seems to need to be within a box of reflective tape, that is 1cm or so wide. That places quite a requirement on the border of the target, on top of the multiple centimeter sized cylinders for reception. How will this targeting work if some of the square is obscured? Might be a problem for a handheld device if the user isn't allowed to hold it. Seems a pretty 'ghetto' marking method for consumer electronics, that is not IMO going to fly in any real product. The transmitter following the receiver was shown to have a short lag each time it moved - is the beam still on and insonifying something else during that time, or switched off? 

Last question on tracking - is there any feedback so the transmitter knows how much power to send, or is it just full on, all the time? If there is, what's the communication method?

While the transmitter isn't shown in the video, in the background a video seems to indicate they are using transmitters like the ones shown last year. Shown in the upper picture next to an iPad, and the lower one marked as "March 2018".

That looks to be about 16 by 16 Murata style devices, each around 1cm in diameter, so 256 elements total and at 145dB (290W/m^2, intensity claimed by uBeam in the past), that would mean around 7.5 Watts acoustic out absolute maximum. If all of that were lighting a, generously say, 100 mW LED then that would mean around 1.4% acoustic to electrical conversion, though when you account for efficiency in the transmitter would drop below 1%. 

They go on to say that they are so awesome for Industrial Internet of Things, low power sensors etc, that's where they are focusing and save everyone the frustration of changing batteries - for the addition of white tape and large receivers around all the sensors, a power bill going up 100x, and only one lunchbox sized transmitter per few sensors. Given that Powercast already sell a wireless power system that will work at up to 80ft, with regulatory approval, can charge at up to the mW level, and does not need large receivers or reflective tape, I'm not quite sure what the value proposition here is.

A 'more robust' version of the demo was being given in their suite at the Venetian, so apparently this demo was not robust. They'd power cameras and sensors, but no mention of phones, which indicates to me that they can't reliably get 500mW to 1W at a phone sized receiver, as generally that's the minimum needed to even start charging a phone.

Safety

This bit was the part that really got me wound up. People are free to do what they want with their own money but safety is where you don't get to screw around. Allow me to rebutt the argument made here that this system is perfectly safe:

All acoustic energy bounces off the skin This is true, you get around 99.9% reflection from bare skin into the air - however - once there is hair on that skin, then acoustic losses go up, and that acoustic energy is converted to heat. Put enough energy there, and it can heat up a lot. Some papers report that mice and rabbits can die from exposure to sound at 145dB and up. From "Effects of Ultrasonic Noise on the Human Body—A Bibliographic Review" 

"According to Allen, Rudnik and Frings, a mouse dies from overheating after 10 s to 3 min of exposure to a signal of 20 kHz and level of 160 dB [10]. According to Danner, a lethal level for signals of 18–20 kHz for an unshaven mouse were 144 dB and for a shaven mouse 155 dB [21]. Acton obtained similar results and extended studies to larger animals such as guinea pigs and rabbits [22]."

Now at 145dB temp rises can be small, but cumulative, and consumer devices have to cover edge cases like babies, drunks, ill people who can't move, pets locked in cages etc. When you sell millions of devices to consumers, who do things they shouldn't, your system has to be foolproof.

I'm wondering, if it is pointed at the ear canal, does it work differently in there, where there is a very sound sensitive part? Any sub-harmonic generation possibilities, where a lower frequency that what is transmitted is heard? (For example, a subharmonics of 40 kHz are 20 and 10 kHz, both in hearing range of some to pretty much everyone). I'm sure that was studied too.

We've had 3rd parties evaluate our system over 18 months and it's completely safe including for pets There is a lack of evidence in the literature that ultrasound in the environment at these amplitudes is safe, and if they have it, this would be a landmark paper that would be massively valuable around the world. I'm sure there was a scientifically controlled experiment, repeated multiple times for statistical significance, followed over years, and if people were involved (which to study human hearing or skin response there would have to be) there will be an ethics review somewhere too? 

Please write a paper, for peer review, and all critics will be silenced. So come on uBeam, release the study - it doesn't have to reveal anything about your system, or technology, just the effect of sound. There's no commercial reason to keep it hidden, no competitors that will steal a march on you with it. All it can do is benefit you - so release it. Or is it not quite that good?

Tim Leighton from University of Southampton did the most comprehensive study to date on the effects of ultrasound in the air I know of. You can read it here, and it is not at all as confident regarding safety effects, especially long term.

The beam is controlled and directed If the wavelength of sound is smaller than the pitch between the transmitting elements, there will be what are known as 'grating lobes' where energy is sent in directions in addition to the desired beam. Given the spacing seems to be around 1cm, and at uBeam stated frequencies the wavelength is smaller than this, there will almost certainly be grating lobes. If so, how many people walking past that demo were getting insonified? That is pretty appalling to me - that members of the public, without their knowledge, could be subjected to unknown sound levels that may or may not have regulatory approval. Which brings me to the next part of safety.

Regulatory. OSHA used to have a limit of 145 dB for sound above 40kHz in air (US only), however a look at the rules today appears to show that it's a flat 115 dB, 1000x less power. Other countries are all in the 110 to 115 dB range. The FDA requires approval for all radiation emitting products, while UL 61010-1, Section 12.5.2 "Protection against... ultrasonic pressure" says "the ultrasonic pressure shall not exceed 110 dB above the reference pressure value of 20 microPa for frequencies between 20 kHz and 100 kHz"

So it seems that OSHA, the FDA, and UL all require much more stringent safety than simply blasting 145 dB around. This will be answered in that 3rd party set of tests, right? Again, no problem releasing this as there was a claim in the Oct 2017 fundraising deck that uBeam was "legally approved by FCC/FDA".

When it comes to regulatory and safety the burden is not on the regulator to prove it is dangerous, it is up to the proposer to prove that it is safe.

Engineering Ethical Considerations

Being an engineer isn't just about doing calculations and building things. There's a responsibility to the public and the world at large on what, and how, we work. The IEEE is the world's largest engineering professional body, and have a set of ethical rules, they can be found here, and the first one among them is:

"to hold paramount the safety, health, and welfare of the public, to strive to comply with ethical design and sustainable development practices, and to disclose promptly factors that might endanger the public or the environment"

So sell anything you like, fools and their money are soon to be parted, but prove that it is safe and don't ever put the health and welfare of people at risk. If you do, don't call yourself an engineer.

Update Jan 19th: Confirmed that it is the white box transmitter.

Update Jan 24th: following some conversations with those who got to see the private suite demo:

The transmitter was on a motorized rotational stage, turning an estimated +/- 45 degrees to steer the beam. If that's the case, I do not understand why they bother with individual elements and a phased array - just get a focused bowl arrangement and steer mechanically, it would be simpler and much cheaper. The CES floor show demo seemed to show phased array operation, so perhaps there's a very limited steering angle and gross motions need mechanical steering?

There were items taped to the wall, on the door etc to show charging of items like "smart locks", however the device itself never charged, it was always an LED that lit up to indicate power was being delivered. That means it could be as low as around 20 mW received.

Those who held the next generation transducers seemed to think them roughly the same lateral dimensions as the Muratas, perhaps a bit thinner, but nowhere close to the "4x smaller area, 100x thinner" listed in the Oct 17 fundraising. They did say that the demo was being done with COTS devices.

Generally the view was that the presentation materials were not particularly professional. Given what they showed, it seems they booked a slot at CES prematurely, I have to think this hurt more than helped - but maybe I'm just a dumb engineer.