 Thank you Hi everyone, so I'm James. I'm actually based in the valley. I'm back in Singapore for Chinese New Year So it just happens that it's a coincidence that this high-quality session is like at the right time, right? So finally I can you know be here give a short presentation about Something that I've been working on for about a past like a half a year or so Okay, so the talking of it is a display wireless power delivery for indoor low-power devices Now later on you'll probably understand why it is only for low-power devices indoors Okay, so the back story how many of you here actually have home automation and raise your hands Okay, so you have like many sensors placed around the house doors windows motion sensors today all use batteries Well, generally if you know you set up a whole automation When you run away a house right most of your things will be wired in so that's fine But for let's say you are just moving in you're renting a place and you have home automation You install these devices which are battery operated All right, so typical home in the states, right? I have a smoke detector some monosan sensors all these kind of door sensors, right? And If we look at other actually operated things Xbox PlayStation 4 Why is people mouse? Some you use some door camera, right? Echo TV remote party vehicle TV all kinds of set up boxes just because we get them there Yeah, so we have all these kinds of things over there and we realize that you know The one one the key pain points is that they use all kinds of different types of batteries And I don't stock batteries at home by eat that I go to Amazon I press a button and everything's ready, right? So they have double a battery triple a battery is my old cells using smoke detectors I don't know the state just loves to use my book sells We're looking CR20123 a batteries which are these little cells that go into the Door or window sensor is not common. And of course you got your pretty common of CR2032 in some little remotes, right? so this is a Approximately run down their capacity and typically these cells or other your devices have A typical runtime of maybe half a year one year or in some for the Amazon fire TV remote control is like one month so in the typical time you go home if I Count all my Virtual operating devices. I actually have more than 30 of them, right? So when I first started off Setting up my home automation system. You know, I was really happy It's like always so cool I saw a sensor here a sensor there on the windows doors and a little bit like near reaching the one year mark The battery started to fail right so on my whole automation panel I start seeing them go and disappearing one by one, right? And I have to go and replace a battery. So okay fine. I bought a couple of batteries I replaced one of them two of them when you went down to like the 22nd 23rd device You are really tired and it is not just one days over a period of time because batteries don't You know run out of power at the same time, right? It's just across the entire time and you know, you just keep distracting you your smoke alarm when you run out of power It wakes you up in the middle of night beeping, right? How pleasant that and typically if you don't remember where your devices are Especially the sensors on doors and windows that are kind of like obscure places you would not know It is down. You don't remember where it is. You don't know what kind of battery users So you open a check you realize that you don't have it in stock you order it you come back a day later So it's just um, yeah Really tedious or if I had to do this my whole list will fill off Things to do like, you know by detergent replace batteries these kind of stuff So when I told you that there's no longer need to replace batteries anymore or I hope Right. So one of the things I thought was you know, what if I could have a beacon in the ceiling, right? Which they can charge all the devices in my room without me carrying them where it is Of course, there's some exceptions if I, you know, put it in the lab box or the fire leakage. It wouldn't charge You'll see why later Okay, so look at mine, right? They're actually a couple of existing solutions. If any of you need to see as early this year, January So at CES, there are a couple of wireless energy wireless power startups, right? There's a energies why should see You've been a spin around for a couple of years But seriously, I mean you just move about it. You any like this is major or something You know why it's kind of silly, but yeah, so in general the key idea is that many of these wireless energy Technologies, right? They have a transmitter. They mean radio frequency Radio energy, right across the distance. It could be centimeters. It could be feet so We could that mind right now, you know, I wonder like this is really dangerous or You know, what what implications of like flying the room with radio radio energy, right? One of the things to note is that many of these devices they operate within Some power limits so over here, right? We see that the there's a little power limit Primary because I'm using that frequency. It's about one watt, right? And just keep in mind that we must have a Yeah, we must conform to these kinds of limitations, right? And many of these Companies they also have to be strictly these kinds of limitations Yeah, so basically don't have infinite power. You have a limited power budget Okay, so there are a couple of couple of different types of antennas used to regulate radio frequency Typically your Wi-Fi router, right? The Wi-Fi router would regulate radio frequency all direction This is typical of antenna radiation pattern, right? But remember that we have a limited power budget So why do you want to waste energy radiating radio frequency all directions? So another idea is that okay, why don't we try to shape This radio energy and push it towards a specific direction Right, so this called mean forming is used in modern Wi-Fi routers actually because we use it for power delivery But you realize that you're actually using radio energy to shape The radiation pattern of radio energy which is you're still kind of like wasting the power budget for that I added the links at the bottom, so you know if you guys like visit the slides after the talk You can go and take a look at what I was reading So my approach I primarily use a directional antenna Okay, I wanted to put the entire power budget in a very narrow beam To point it towards the target device that intends to receive the power, right? So a key benefit of this is that I don't have real energy, you know radiating all around It's very directional I did some preliminary tests walking around the meter to see that it actually radiates the whole pattern You can just get these antennas I got mine from eBay for about like 30 bucks So they have a very narrow beam with so you can imagine all these power all this power, right? In that specific radiating in a specific direction About it looks large and some I think some people in the community they've made out antennas So you could make it lighter and cheaper if you do it yourself So the general approach is that we want to scan the room for Devices to be charged and if we find devices to be charged to charge them, right? Typically, you know, if it's less than 60% we'll try to charge it and we just like 180% we stop charging So yes, these devices still have a battery Well, they need a battery to store charge But the good thing is that I just hope that we have we no longer need to replace them Okay, and of course, there's an exception if we can't charge the device faster than is being drained It should inform the user usually in these cases No, the battery is like 40 or spoiled But you know, hopefully this will take a couple of years before it kicks in, right? So so the typical beacon it should just like running energy It wakes up the device and device feedback to me how much charge it has and this happens During a scanning phase, which I'll demonstrate later Okay, so not so the intent is very directional, right? Which means that we have to move it around in order in all angles, right in order to cover a suppose like assume like the 10 to 15 feet radius and It should yeah, it should be able to point to various parts of the room So in my prototype, I basically grab a couple of servo motors dynamic So anyone use dynamics or servos here in your robotics projects or No, okay And there's another servo called super 200, which I got from AliExpress. It is really crusty Chinesium stuff. So if you want to read about my difficulties of getting it to Actually do its work. You can click on the link below in future So yeah, my demo is over there I didn't want to put it here because it's too far from everyone It does a power socket there. We could plug it in and I can show you how it lights up an LED wirelessly, you know But in general this just spins around Right, the model is that I will actually spin move the antenna around It would don't do like a 360 degree scan around the room for wireless devices and also change the face Because antenna is a polarization, right? So you change the face of the antenna as it goes around So this is just a quick hack done Wooden prototype just to allow me to like collect some data. Okay, and the receiver I took quite a while to get this but basically we have to have an antenna. It's a very naive design I'm quite sure like if there's any RF experts or engineers here, you can call it a better job But it's a simple die-hole, you know, I did some impedance matching hooked them to a receiver the receiver basically rectifies the RF Energy right to get the DC voltage out of it and we feed the DC voltage into the charge controller Yeah, I've a list of materials in the next slide. Yeah, but the key idea here is that it's very important to get your antenna matching rights Or else you would not have a good good power efficiency And of course, it's not a design. It can be improved the last Measurement I told at about 15 10 15 fit is about less than 10 million watts of receive power Now may seem really little but in the grand scheme of things when you're talking about Recharging low power sensors, IoT devices around the home 10 million watts can go a long way and think about it You have like 24 hours a day for it to run, right? It's not like, you know, you have a short amount of time But you have the whole day for this to slowly charge up the devices So quick run of key components transmitter, right? I just like the transmitter the highest possible power within safe limits Well, I wouldn't want to You know, like be exposed to read too much radiation like a microwave oven So a large antenna we try to look for the Antenna with the most narrow beam with you can find Currently, I think it's something around 23 dpi I try to go higher a smaller beam with means the antenna needs to be larger, which I don't want something that's too large Yeah, the receiver antenna probably have to run through some like the simulations in like HMSS All those kinds of like RM simulation software to get a good Yeah, a good design The RFC component is pretty important because you can't simply drop in any random dial You fine, right? Yes be something, you know, that has like How would I say a very little power loss? Shop key. So I actually this is two models, right? One is the one SS 106 Hitachi, I'm gonna test this. I just Google it online. Someone use that to light an LED. That's great I use the RFD1028 It's actually like a early product from some Other microwave engineer in the States. Yeah A charge controller is a pretty mainstream be view 25504 a very low loss works up to like 0.1 or 0.2 volts. So that's great. You can charge up the battery like a lithium cell or 1.2 volts nickel metal hydrate so Yeah, and of course your storage Preferably we choose a small cell because small cell results in faster voltage rise over time and you charge it up So results in better useful voltage for your system. So some of the problems when I'm occurring I was testing So I didn't actually test this myself by someone to help me, right? It's getting all the cheap Transmitter which I got from China, right has a pretty bad. How would I say? It's a pretty transparent pretty white man Is he so so naturally my antenna is narrow band and he only captures my part of power many of these other power Transmissions and other frequency is lost Yeah, so, you know, if you design your own transmitter definitely you can improve on this part You could like have the entire one or two watts in demand that you want so Now you may think like, okay, is it hard to Visualize know when when you move the antenna around you scan around the room Okay, is it hard to know where the device is in real space? My point is that it is not hard because I read I really done this before in the previous project I think my final year project in NUS Where I kind of like tilt the antenna and scan the physical space And you know, we just pull the graph and you can easily see that there are clearly points where A certain orientation results in better signal strength than other positions Yeah, at least educated up from another experiment, but it's a similar idea But sometimes, you know, you may encounter because like indoors, right? You may have like reflections along the walls or a metal cabinet or other things, right? So you may find that two positions or two orientations of your transmitter will result in A good signal being received to receive Yeah, so with this current prototype You know, which are probably where I find but I'm actually a software engineer by day and by night I'm like taking all kinds of stuff tearing down China parts. But yeah, I think I probably collect more data Work on the battery transmitter communication improve the antenna design actually needs more work because currently it's very polarized Yeah, so I think that's all for now where I got a Yeah And just in case anyone is interested in related works I used to do some research in antenna optimization physically moving an omnidirectional antenna around But those were undergrad days Yeah Anyone has any questions? Yes No, I'm not actually using digital beamforming Primary because I don't firstly is actually hard, right? I need to have like many amplifiers to no change the face of the beam face of every antenna multiple antennas And as as I demonstrated earlier, you know, there's a lot of energy loss In beamforming I need to use a lot of rm energy to form the beam in the first place No, no, no, I'm mechanically steering the antenna around Yes, but they are how would I say? Is the antenna that has a narrow beam width because of those nine antennas the beam width is narrow and it's already designed that way Yeah