WEBVTT 00:00.000 --> 00:07.000 Let's see if I'm sure then. 00:07.000 --> 00:12.000 So welcome, and thanks for being here. 00:12.000 --> 00:18.000 So this talk is about a very, do it yourself project. 00:18.000 --> 00:26.000 And this is how I made some smartness into a done EV charger for, 00:26.000 --> 00:31.000 well, for open source and for being viable for me 00:31.000 --> 00:35.000 to charge a car at my parents. 00:35.000 --> 00:37.000 Okay, I'll try to speak louder. 00:37.000 --> 00:39.000 My voice is really getting a hard time. 00:39.000 --> 00:41.000 Thanks. 00:41.000 --> 00:45.000 So I'm currently a kickhouse developer at the, 00:45.000 --> 00:49.000 a tiny bit, which is a company we do with data analytics. 00:49.000 --> 00:52.000 I presented a track like founded this. 00:53.000 --> 00:59.000 I presented on 1, 2019 about privacy and showing who is tracking us 00:59.000 --> 01:00.000 through the internet. 01:00.000 --> 01:04.000 And I'm also a hip-olycerly user group that exists in a cornea. 01:04.000 --> 01:07.000 And we are organizing also a free software conference there, 01:07.000 --> 01:09.000 and CFP is open. 01:09.000 --> 01:18.000 So, okay, so I made my own way of interacting with a car for, 01:18.000 --> 01:19.000 for charging. 01:19.000 --> 01:22.000 And first of all, the reason why I did this is this was a site project, 01:22.000 --> 01:24.000 and I wanted to learn Rust. 01:24.000 --> 01:30.000 So I ended up making this as an excuse as a project to learn the language. 01:30.000 --> 01:34.000 And I talked yesterday at the Rust that room about how I did it, 01:34.000 --> 01:37.000 the talk of today is why I did it. 01:37.000 --> 01:40.000 So why? 01:40.000 --> 01:46.000 Well, I now own an EV and charging cheaper than the public pricing 01:47.000 --> 01:52.000 and my parents lived in the countryside where there's a very poor electric power. 01:52.000 --> 01:55.000 There's less than 2.2 kilowatts peak available. 01:55.000 --> 02:00.000 And most of the chargers that are currently available do a kind of, 02:00.000 --> 02:02.000 this smart grid balancing and so on, 02:02.000 --> 02:06.000 but they are only interested in high power output for 10 cars or 20 cars, 02:06.000 --> 02:08.000 and this is not my use case. 02:08.000 --> 02:11.000 These balancing chargers exist, 02:11.000 --> 02:14.000 but first of all, they usually require physical wiring, 02:14.000 --> 02:18.000 which is cumbersome because when you have a car that is like 15 meters away from 02:18.000 --> 02:21.000 your circuit breaker in a different area, 02:21.000 --> 02:27.000 then you would have to do a different groundwork to put a wire. 02:27.000 --> 02:31.000 Just a tiny little wire that only carries like one volt to check 02:31.000 --> 02:34.000 what's the amperage of the breaker. 02:34.000 --> 02:38.000 So it's really expensive and really not convenient for this to do. 02:38.000 --> 02:40.000 This is kind of displaced. 02:40.000 --> 02:42.000 So it doesn't really make a lot of sense to like, 02:42.000 --> 02:47.000 all through like 15 meters away to just go to the other side and charge the car there. 02:47.000 --> 02:51.000 Because we already had, we already had AC there. 02:51.000 --> 02:55.000 We just didn't have that one wire. 02:55.000 --> 02:58.000 So the thing is also that cost matters, 02:58.000 --> 03:03.000 because part of the sighting to purchase an EV was the cheap electricity 03:03.000 --> 03:05.000 and the convenience, like 10 years across, 03:05.000 --> 03:08.000 that it would be cheaper than running a gas car. 03:08.000 --> 03:11.000 So given that this is also decreasing the preheating acid, 03:11.000 --> 03:15.000 it doesn't make sense to spend more on chargers and other infrastructure 03:15.000 --> 03:21.000 than it's needed because then you're just moving the milestones 03:21.000 --> 03:24.000 and it doesn't make sense economically. 03:24.000 --> 03:26.000 As I said, this is expensive. 03:26.000 --> 03:28.000 Should they? 03:28.000 --> 03:31.000 Well, the tech is very complex right now. 03:31.000 --> 03:35.000 Many of these chargers that are quite intelligent 03:35.000 --> 03:38.000 that they have hardware inside, like a Raspberry Pi or something like that, 03:38.000 --> 03:40.000 which is already quite expensive on itself. 03:40.000 --> 03:44.000 It's like 50-60 euros, just the device. 03:44.000 --> 03:48.000 And if the goal is saving money, then this is my economics. 03:48.000 --> 03:54.000 If I charge at home, I would be charging at around 18 cents of an euro, 03:54.000 --> 03:59.000 particularly what I learned that I have already been charging at the average of 34, 03:59.000 --> 04:03.000 which means that if I visited my parents twice a year twice a month for a year, 04:03.000 --> 04:06.000 I get to save at most 80 euros. 04:06.000 --> 04:10.000 So, there's a point where it doesn't make sense to purchase a charger. 04:10.000 --> 04:14.000 Oh, it's just less expensive for me to just charge a type, 04:14.000 --> 04:18.000 which doesn't make a lot of sense for the charging industry. 04:18.000 --> 04:22.000 So, as I said, 04:22.000 --> 04:26.000 the Turkey solutions are expensive and also the open solutions that exist 04:26.000 --> 04:29.000 on opening VSC and so on are also very expensive, 04:29.000 --> 04:35.000 because they use computing hardware that is quite expensive to have in an embedded device. 04:35.000 --> 04:42.000 And there are reasons for this of course, but can we do more with less than that? 04:42.000 --> 04:49.000 So, what I did is I tried to research for what I think is the sensible way of going through this, 04:49.000 --> 04:52.000 trying to see what is already existing. 04:52.000 --> 04:54.000 How could I make it? 04:54.000 --> 04:57.000 What are the standard simple things in charging? 04:57.000 --> 05:00.000 What are the projects of existing open source that implement this? 05:00.000 --> 05:06.000 And maybe develop something that doesn't anything that is missing. 05:06.000 --> 05:11.000 And keeping mind that what I want to do is I wanted to just be able to charge the car 05:11.000 --> 05:13.000 at the maximum rate possible, 05:13.000 --> 05:18.000 but keeping in mind that things like having running water and having rich open 05:18.000 --> 05:23.000 and that the secret breaker doesn't break is the main point here. 05:23.000 --> 05:28.000 And also maximum, the car charge rate, 05:28.000 --> 05:32.000 because I could just charge out one amp, but that wouldn't take me there here long. 05:32.000 --> 05:35.000 So, there's two standards. 05:35.000 --> 05:40.000 How many of you know about how EVSE is charging what are the standards around this? 05:40.000 --> 05:42.000 Okay. 05:42.000 --> 05:46.000 Okay, for the majority of the room doesn't so let's talk a bit about this. 05:46.000 --> 05:50.000 There's basically two main standards about going this way. 05:50.000 --> 05:53.000 Here's the letter one, the IECU 6181. 05:53.000 --> 06:00.000 This is electric signaling that uses pulse with modulated signal, 06:00.000 --> 06:05.000 where depending on how much of the pulse is full, 06:05.000 --> 06:08.000 then this signals how much of an amp you are requesting. 06:08.000 --> 06:11.000 So, for each one percent of the cycle that you had, 06:11.000 --> 06:16.000 this is 0.6 amps, this was based on an American standard before. 06:16.000 --> 06:18.000 And this is very cheap to implement. 06:18.000 --> 06:22.000 You don't even need an integrated circuit to do this. 06:22.000 --> 06:24.000 You could do it with a 555 timer, 06:24.000 --> 06:29.000 but you can also do this with an SP32 with anything that has the capacity to output one volt 06:29.000 --> 06:35.000 and then have a couple of resistors and the capacitors to do the rest of the wiring. 06:35.000 --> 06:38.000 This is very, very easy to do. 06:38.000 --> 06:43.000 The bad thing about this is that since 0.6 amps is the, 06:44.000 --> 06:48.000 it's the one percent and the minimum if it's a 10 percent on the standard, 06:48.000 --> 06:53.000 you can't really tell the car to charge at less than 6 amps. 06:53.000 --> 06:56.000 And in my case, I had between 0 and 10 amps. 06:56.000 --> 06:59.000 So if I only have from 6 to 10, 06:59.000 --> 07:03.000 then for half the time where anything is being run at home, 07:03.000 --> 07:07.000 the car would just be sitting idle, which is not convenient. 07:07.000 --> 07:13.000 Then there's the more sophisticated ISOIC-15118 standard, 07:13.000 --> 07:16.000 which is a lot more convenient to implement, 07:16.000 --> 07:18.000 because this is packet based. 07:18.000 --> 07:23.000 So on both of these things, 07:23.000 --> 07:25.000 what basically what you have, 07:25.000 --> 07:29.000 if you have ever seen one of these chargers at the street or anywhere, 07:29.000 --> 07:32.000 you will see that there's five connectors. 07:32.000 --> 07:36.000 One of them is called the control pilot and what it does is, 07:36.000 --> 07:39.000 this is that one and earth is the only path 07:39.000 --> 07:42.000 where the charger communicates to the car. 07:42.000 --> 07:45.000 So this ISO is standard. 07:45.000 --> 07:46.000 Well, the people that have the same standard, 07:46.000 --> 07:48.000 just uses that wire to make up all that, 07:48.000 --> 07:52.000 with modulated signal, which is very easy. 07:52.000 --> 07:56.000 But the ISO, what it does is, over that signal, 07:56.000 --> 07:59.000 it has to send Ethernet packets, 07:59.000 --> 08:04.000 which requires you to use something that is called 08:04.000 --> 08:08.000 a PLC. 08:08.000 --> 08:12.000 You may have seen this, for example, when you use, 08:12.000 --> 08:16.000 there's these solutions where you put two devices on the AC, 08:16.000 --> 08:19.000 directly, and then you put, you have Ethernet over there, 08:19.000 --> 08:22.000 that's using this sort of same kind of communication. 08:22.000 --> 08:25.000 The thing is, these modules are quite expensive. 08:25.000 --> 08:29.000 I didn't see them online to be able to purchase them for less than 60 euros. 08:29.000 --> 08:33.000 Maybe there's, I mean, probably for integrators, 08:33.000 --> 08:35.000 this is cheaper, but for me, it's an individual, 08:35.000 --> 08:39.000 I couldn't even participate of getting this in an open source, 08:39.000 --> 08:42.000 how we can think for less than that. 08:42.000 --> 08:47.000 And I would also have to implement over this Ethernet thing, 08:47.000 --> 08:51.000 also the actual signaling with the car, which is XML. 08:51.000 --> 08:55.000 So you will either have to do all of the XML signaling 08:55.000 --> 08:57.000 in some embedded kinds of devices, like, 08:57.000 --> 08:59.000 on ESP32 or something like this, 08:59.000 --> 09:01.000 but the things that are already built, 09:01.000 --> 09:04.000 like in opening the EVSE, and some other projects, 09:04.000 --> 09:08.000 are using currently full Linux OS with Python 09:08.000 --> 09:13.000 to run the XML parser basically, 09:13.000 --> 09:15.000 which means that there's also a lot of overhead 09:15.000 --> 09:20.000 in terms of, like, the kind of computer devices that we are using for this. 09:20.000 --> 09:22.000 So maybe there's a third way. 09:22.000 --> 09:25.000 I used something that is not standard here, 09:25.000 --> 09:27.000 which is the car vendor API. 09:27.000 --> 09:30.000 This is for different providers. 09:30.000 --> 09:34.000 This is the one I'm using, where you have an API, 09:34.000 --> 09:39.000 that using the own car 5G car, 09:39.000 --> 09:42.000 you can ask it how much it should charge. 09:42.000 --> 09:46.000 This is possible in AC, because when you charge car in AC, 09:46.000 --> 09:48.000 it's the onboard charging in the car, 09:48.000 --> 09:52.000 the one who knows how much it can charge. 09:52.000 --> 09:55.000 So you can request it directly to that. 09:56.000 --> 09:57.000 So what did I do? 09:57.000 --> 09:59.000 I first bought the dumbest charger I could, 09:59.000 --> 10:01.000 and the cheapest one I could find 10:01.000 --> 10:03.000 to write to understand how this works. 10:03.000 --> 10:05.000 Step two, this assembly. 10:05.000 --> 10:09.000 To check whether the cheapest thing has enough 10:09.000 --> 10:12.000 of what it needs to, to not break apart 10:12.000 --> 10:14.000 and catch up in flames. 10:14.000 --> 10:15.000 And it does. 10:15.000 --> 10:18.000 Basically, the one thing that is the most expensive, 10:18.000 --> 10:21.000 the thing that is most expensive here, 10:21.000 --> 10:23.000 if I wanted to do this on my own, 10:23.000 --> 10:25.000 which I also thought about doing, 10:25.000 --> 10:28.000 is the big wires here. 10:28.000 --> 10:30.000 I'll step away from the camera for a moment. 10:30.000 --> 10:33.000 These three wires, the five meters of cable, 10:33.000 --> 10:35.000 are the most expensive thing on a charger. 10:35.000 --> 10:36.000 Just the five meters of wire, 10:36.000 --> 10:39.000 can run you maybe like 40, 50 years of the 60, 10:39.000 --> 10:42.000 that this costs, is basically copper. 10:42.000 --> 10:45.000 Copper is really expensive nowadays. 10:45.000 --> 10:47.000 The rest of it is very cheap electronics. 10:47.000 --> 10:51.000 You have a couple of relays to switch on 10:51.000 --> 10:53.000 and off the GAC current. 10:53.000 --> 10:57.000 And then, well, you can see there's not much. 10:57.000 --> 11:01.000 The one thing that is maybe a bit interesting to see here, 11:01.000 --> 11:04.000 is that there's this small chip, 11:04.000 --> 11:06.000 this is a, 11:06.000 --> 11:08.000 he got the bias, 11:08.000 --> 11:10.000 A Rm chip, this is a small, 11:10.000 --> 11:13.000 A Rm chip that is doing the post, 11:13.000 --> 11:16.000 with more related thing. 11:16.000 --> 11:19.000 I could change this and program it to do whatever I wanted, 11:19.000 --> 11:22.000 but I couldn't find a new sheet easily online to flash it. 11:22.000 --> 11:23.000 So it was, 11:23.000 --> 11:25.000 if I wanted to do this with, 11:25.000 --> 11:28.000 with the IEC standard, 11:28.000 --> 11:30.000 I could just remove this wire, 11:30.000 --> 11:32.000 this is the control pilot that I mentioned, 11:32.000 --> 11:35.000 and run this through my own microcontroller, 11:35.000 --> 11:36.000 and then talk to the car, 11:36.000 --> 11:39.000 and then put another wire here so that the charger knows 11:39.000 --> 11:41.000 that I'm talking to something that is like the car, 11:41.000 --> 11:43.000 so that the charger knows I'm talking to something, 11:43.000 --> 11:44.000 the car knows I'm talking to something, 11:44.000 --> 11:46.000 and I put something in the middle to make things work. 11:46.000 --> 11:47.000 This is what I thought, 11:47.000 --> 11:49.000 I had in mind at the first time, 11:49.000 --> 11:51.000 but then I saw the car API, 11:51.000 --> 11:53.000 it was very much easier to do that way, 11:53.000 --> 11:55.000 so I started in that, 11:55.000 --> 11:58.000 but if we all wanted to do something with this, 11:58.000 --> 12:00.000 this is a very possible and easy. 12:00.000 --> 12:03.000 This is basically a Cortex M3, 12:03.000 --> 12:05.000 with 72 megahertz, 12:05.000 --> 12:08.000 it's a really, really small microcontroller. 12:08.000 --> 12:13.000 So in order for me to be able to do this, 12:13.000 --> 12:16.000 smart and be able to do a grid balancing, 12:16.000 --> 12:18.000 I would have to build a wattmeter, 12:18.000 --> 12:20.000 and I wanted to do this wirelessly, 12:20.000 --> 12:24.000 so that I didn't have to run wires over the countryside, 12:24.000 --> 12:28.000 to tell the car a plus something to let 12:28.000 --> 12:30.000 the car know how many I'm still request. 12:30.000 --> 12:32.000 So this wattmeter, 12:32.000 --> 12:36.000 instead of all of the money that we would cost to other things, 12:36.000 --> 12:39.000 I, well, you can see, 12:39.000 --> 12:41.000 this is really cheap to do. 12:41.000 --> 12:44.000 The current transformer is just, 12:44.000 --> 12:47.000 I'm not sure if I don't think they have here, 12:47.000 --> 12:50.000 the picture of the city anyway. 12:50.000 --> 12:53.000 This is basically, 12:53.000 --> 12:55.000 the current transformer is something that you, 12:55.000 --> 12:56.000 yeah, it's a coil that you, 12:56.000 --> 12:57.000 you put one wire over it, 12:57.000 --> 12:59.000 and it will measure in both, 12:59.000 --> 13:02.000 how many amps are being passed through the wire, 13:02.000 --> 13:05.000 really small electronics device, 13:05.000 --> 13:07.000 not even electronics. 13:07.000 --> 13:11.000 And then a power source to power this from the, 13:11.000 --> 13:13.000 from the grid. 13:14.000 --> 13:15.000 And then, yeah, 13:15.000 --> 13:18.000 something to the car, which is an API. 13:18.000 --> 13:22.000 For doing this with the, 13:22.000 --> 13:25.000 the SP32, I used Rust, 13:25.000 --> 13:28.000 because you can see my talk just today, 13:28.000 --> 13:31.000 because it just works, 13:31.000 --> 13:32.000 it's easier to build this, 13:32.000 --> 13:34.000 but then with the C++ SDK in my opinion, 13:34.000 --> 13:36.000 if you want to do something with this, 13:36.000 --> 13:38.000 I highly recommend it. 13:38.000 --> 13:40.000 And then, sorry, 13:40.000 --> 13:42.000 and then, 13:42.000 --> 13:43.000 here it is. 13:43.000 --> 13:44.000 Just, 13:44.000 --> 13:47.000 well, this is basically the current transformer, 13:47.000 --> 13:49.000 this is running to AC directly, 13:49.000 --> 13:50.000 and then, 13:50.000 --> 13:52.000 this is the current transformer, 13:52.000 --> 13:54.000 you just put one of the wires over here, 13:54.000 --> 13:56.000 and you will measure how much, 13:56.000 --> 13:59.000 so it will measure one bolt over this, 13:59.000 --> 14:00.000 over this line, 14:00.000 --> 14:01.000 if there's 30 amps here, 14:01.000 --> 14:03.000 and otherwise it's just linearly, 14:03.000 --> 14:05.000 whatever it is. 14:07.000 --> 14:08.000 So, I, 14:08.000 --> 14:09.000 oops, 14:09.000 --> 14:10.000 okay. 14:11.000 --> 14:13.000 So now, I could do this manually, 14:13.000 --> 14:15.000 I could just go and curl every, 14:15.000 --> 14:17.000 a couple of seconds for the, 14:17.000 --> 14:19.000 the website that I put on this device, 14:19.000 --> 14:21.000 then, 14:21.000 --> 14:22.000 step, 14:22.000 --> 14:23.000 I could automate it, 14:23.000 --> 14:26.000 and set the charging amps to the car, 14:26.000 --> 14:27.000 every few times, 14:27.000 --> 14:30.000 then I could put this in a more sophisticated script, 14:30.000 --> 14:32.000 but still running from my laptop, 14:32.000 --> 14:33.000 this is not ideal, 14:33.000 --> 14:34.000 because, well, 14:34.000 --> 14:36.000 all of these requirements to have a laptop here, 14:36.000 --> 14:37.000 have LAN access, 14:37.000 --> 14:38.000 and, 14:38.000 --> 14:39.000 well, I just have to remember to do this. 14:40.000 --> 14:41.000 So, 14:41.000 --> 14:42.000 I think, 14:42.000 --> 14:44.000 tap to the car automatically. 14:44.000 --> 14:45.000 Basically, 14:45.000 --> 14:46.000 these things, 14:46.000 --> 14:47.000 so, 14:47.000 --> 14:48.000 I used, 14:48.000 --> 14:50.000 I used a backend with, 14:50.000 --> 14:51.000 with Rust, 14:51.000 --> 14:52.000 and, 14:52.000 --> 14:54.000 for this, 14:54.000 --> 14:57.000 I used Rocket, 14:57.000 --> 14:59.000 because of several reasons, 14:59.000 --> 15:00.000 regarding Rust, 15:00.000 --> 15:01.000 basically, 15:01.000 --> 15:03.000 to ensure that the, 15:03.000 --> 15:05.000 the API requests were valid, 15:05.000 --> 15:07.000 with the, 15:07.000 --> 15:08.000 with the database token, 15:08.000 --> 15:09.000 and so on. 15:09.000 --> 15:10.000 And then, 15:10.000 --> 15:11.000 I even did, 15:11.000 --> 15:12.000 I mapped the vendors, 15:12.000 --> 15:13.000 the, 15:13.000 --> 15:14.000 the energy usage over the, 15:14.000 --> 15:15.000 of the house, 15:15.000 --> 15:16.000 over a period of time. 15:16.000 --> 15:17.000 This was even useful to me, 15:17.000 --> 15:18.000 outside of the, 15:18.000 --> 15:19.000 the thing, 15:19.000 --> 15:20.000 because, 15:20.000 --> 15:21.000 we noticed that, 15:21.000 --> 15:23.000 when, 15:23.000 --> 15:25.000 this, 15:25.000 --> 15:26.000 okay, 15:26.000 --> 15:28.000 this, 15:28.000 --> 15:29.000 at this home, 15:29.000 --> 15:31.000 we have, 15:31.000 --> 15:33.000 our running water, 15:33.000 --> 15:35.000 it comes from, 15:36.000 --> 15:37.000 well, 15:37.000 --> 15:38.000 and, 15:38.000 --> 15:39.000 we have, 15:39.000 --> 15:40.000 water pump to, 15:40.000 --> 15:41.000 to get it through, 15:41.000 --> 15:43.000 and we noticed that we had, 15:43.000 --> 15:45.000 picks of electricity running, 15:45.000 --> 15:46.000 due to the pump, 15:46.000 --> 15:47.000 by just looking, 15:47.000 --> 15:49.000 every few seconds on the, 15:49.000 --> 15:50.000 on the wire, 15:50.000 --> 15:51.000 and we could notice that, 15:51.000 --> 15:52.000 that we, 15:52.000 --> 15:53.000 we should put something to, 15:53.000 --> 15:54.000 to filter this, 15:54.000 --> 15:55.000 this noise on the, 15:55.000 --> 15:56.000 on the, 15:56.000 --> 15:57.000 on the, 15:57.000 --> 15:58.000 like, 15:58.000 --> 15:59.000 to, 15:59.000 --> 16:00.000 anyway, 16:00.000 --> 16:01.000 I didn't like this, 16:01.000 --> 16:02.000 very much in the, 16:02.000 --> 16:03.000 the, 16:03.000 --> 16:04.000 the car vendor API, 16:04.000 --> 16:06.000 I would have preferred to use one of the standards, 16:06.000 --> 16:07.000 but, 16:07.000 --> 16:08.000 unfortunately, 16:08.000 --> 16:09.000 um, 16:09.000 --> 16:10.000 as I said, 16:10.000 --> 16:12.000 there's no current open source implementations of, 16:12.000 --> 16:13.000 uh, 16:13.000 --> 16:14.000 talking via the XML, 16:14.000 --> 16:15.000 uh, 16:15.000 --> 16:16.000 interface with, 16:16.000 --> 16:17.000 with the car, 16:17.000 --> 16:19.000 that doesn't require you to have a full OS, 16:19.000 --> 16:20.000 um, 16:20.000 --> 16:21.000 as, 16:21.000 --> 16:22.000 as such. 16:22.000 --> 16:23.000 So, 16:23.000 --> 16:24.000 I started with this, 16:24.000 --> 16:25.000 because start small, 16:25.000 --> 16:26.000 and then, 16:26.000 --> 16:27.000 go incrementally from, 16:27.000 --> 16:28.000 from here. 16:28.000 --> 16:29.000 Uh, 16:29.000 --> 16:30.000 this is very easy, 16:30.000 --> 16:32.000 and if we wanted to, 16:33.000 --> 16:35.000 be implemented for some other kind of charge, 16:35.000 --> 16:36.000 or some other kind of, 16:36.000 --> 16:37.000 car API, 16:37.000 --> 16:38.000 this would be very easy, 16:38.000 --> 16:40.000 because most of the work is already done, 16:40.000 --> 16:41.000 in a way that is, 16:41.000 --> 16:42.000 um, 16:42.000 --> 16:43.000 uh, 16:43.000 --> 16:44.000 pluggable. 16:44.000 --> 16:45.000 So, you just need to change the, 16:45.000 --> 16:47.000 eV charge handler kind of thing, 16:47.000 --> 16:48.000 um, 16:48.000 --> 16:50.000 which is something that you only have to implement two things, 16:50.000 --> 16:51.000 well, 16:51.000 --> 16:52.000 actually three, 16:52.000 --> 16:53.000 uh, 16:53.000 --> 16:54.000 one is, 16:54.000 --> 16:55.000 uh, 16:55.000 --> 16:56.000 the kind of object that you want to keep, 16:56.000 --> 16:58.000 as state of what you, 16:58.000 --> 16:59.000 uh, 16:59.000 --> 17:00.000 what you have internally, 17:00.000 --> 17:01.000 which can be whatever you want, 17:01.000 --> 17:02.000 um, 17:02.000 --> 17:03.000 and then, 17:03.000 --> 17:04.000 uh, 17:04.000 --> 17:05.000 only two methods. 17:05.000 --> 17:06.000 One of them is, 17:06.000 --> 17:07.000 get the state, 17:07.000 --> 17:08.000 so that we can, 17:08.000 --> 17:09.000 um, 17:09.000 --> 17:10.000 and the state can be whatever you want, 17:10.000 --> 17:11.000 with some restrictions, 17:11.000 --> 17:12.000 that I have on the next slide, 17:12.000 --> 17:13.000 and then, 17:13.000 --> 17:14.000 request, 17:14.000 --> 17:15.000 uh, 17:15.000 --> 17:16.000 something to, 17:16.000 --> 17:18.000 ask the car to charge at the specific rate. 17:18.000 --> 17:19.000 This can be in, 17:19.000 --> 17:20.000 in the charger, 17:20.000 --> 17:21.000 in the car, 17:21.000 --> 17:22.000 whatever it is. 17:22.000 --> 17:23.000 Uh, 17:23.000 --> 17:24.000 and then there's the state, 17:24.000 --> 17:25.000 the state has to implement five things, 17:25.000 --> 17:26.000 uh, 17:26.000 --> 17:27.000 which is whether it's charging or not, 17:27.000 --> 17:28.000 because if it's not, 17:28.000 --> 17:29.000 uh, 17:29.000 --> 17:30.000 we, 17:30.000 --> 17:32.000 and ask repeatedly things, 17:32.000 --> 17:33.000 until it's done, 17:33.000 --> 17:34.000 and then, 17:34.000 --> 17:35.000 uh, 17:35.000 --> 17:36.000 get the current charge, 17:36.000 --> 17:37.000 get the last request items, 17:37.000 --> 17:39.000 because this may not be the same, 17:39.000 --> 17:40.000 because it, 17:40.000 --> 17:41.000 uh, 17:41.000 --> 17:42.000 takes a while for, 17:42.000 --> 17:43.000 for the things to ramp up, 17:43.000 --> 17:44.000 so this should, 17:44.000 --> 17:45.000 uh, 17:45.000 --> 17:46.000 this should, 17:46.000 --> 17:47.000 remember what was requested, 17:47.000 --> 17:48.000 and this should measure actually on the car. 17:48.000 --> 17:49.000 And this is, 17:49.000 --> 17:50.000 uh, 17:50.000 --> 17:51.000 well, 17:51.000 --> 17:52.000 I also got this, 17:52.000 --> 17:53.000 get the car distance to the point, 17:53.000 --> 17:54.000 in kilometers, 17:54.000 --> 17:55.000 because if you have a way to, 17:55.000 --> 17:56.000 notice, 17:56.000 --> 17:57.000 then we kick, 17:57.000 --> 17:59.000 we can just sleep for hours, 17:59.000 --> 18:01.000 if you are 500 kilometers away, 18:01.000 --> 18:02.000 you don't have to ask, 18:02.000 --> 18:03.000 uh, 18:03.000 --> 18:04.000 for the state, 18:04.000 --> 18:05.000 for at least five hours, 18:05.000 --> 18:06.000 uh, 18:06.000 --> 18:08.000 if you're going at legal speed limits. 18:08.000 --> 18:09.000 Um, 18:09.000 --> 18:11.000 this is the interface that, 18:11.000 --> 18:12.000 uh, 18:12.000 --> 18:13.000 I am rendering right now on, 18:13.000 --> 18:14.000 on this, 18:14.000 --> 18:15.000 so every, 18:15.000 --> 18:16.000 uh, 18:16.000 --> 18:17.000 you can see every three to six seconds, 18:17.000 --> 18:18.000 something like that, 18:18.000 --> 18:19.000 uh, 18:19.000 --> 18:20.000 the, 18:20.000 --> 18:21.000 the water meter is pushing, 18:21.000 --> 18:22.000 um, 18:22.000 --> 18:24.000 a packet with the, 18:24.000 --> 18:25.000 with the amps at that point, 18:25.000 --> 18:26.000 and then, 18:26.000 --> 18:27.000 uh, 18:27.000 --> 18:28.000 the picks in red, 18:28.000 --> 18:29.000 and then in, 18:29.000 --> 18:30.000 in blue the, 18:30.000 --> 18:32.000 the average over a few, 18:32.000 --> 18:33.000 a few, 18:33.000 --> 18:34.000 a few, 18:34.000 --> 18:35.000 a few seconds, 18:35.000 --> 18:39.000 uh, 18:39.000 --> 18:40.000 uh, 18:40.000 --> 18:41.000 of the, 18:41.000 --> 18:43.000 of the, 18:43.000 --> 18:44.000 of the, 18:44.000 --> 18:46.000 of the energy usage. 18:46.000 --> 18:47.000 This is, 18:47.000 --> 18:49.000 this was quite fast for me to do, 18:49.000 --> 18:50.000 um, 18:50.000 --> 18:51.000 I only, 18:51.000 --> 18:52.000 I took approximately, 18:52.000 --> 18:54.000 maybe one month of total development time, 18:54.000 --> 18:56.000 um, 18:56.000 --> 18:57.000 better-show charge car from, 18:57.000 --> 18:59.000 from scratch, 18:59.000 --> 19:00.000 Without, 19:00.000 --> 19:01.000 uh, 19:01.000 --> 19:03.000 with a budget of, 19:03.000 --> 19:05.000 seventy years at most, 19:05.000 --> 19:06.000 because this, 19:06.000 --> 19:08.000 The charger was sixty-three, 19:08.000 --> 19:10.000 plus the, 19:10.000 --> 19:12.000 Three dollars of the, 19:12.000 --> 19:14.000 opportunities plus four of the other, 19:14.000 --> 19:15.000 uh, 19:15.000 --> 19:17.000 uh, 19:17.000 --> 19:18.000 so I mean, 19:18.000 --> 19:20.000 Um, 19:20.000 --> 19:23.000 compared to running and open and 19:23.000 --> 19:24.280 VSE which, 19:24.280 --> 19:31.400 a winning, in this case, I wanted to highlight this room because, well, basically, because 19:31.400 --> 19:38.440 there's nothing currently that that does this, I had to do this because I didn't want 19:38.440 --> 19:48.520 to run a full Linux just to talk some XML with with car or just to send a few packets 19:48.520 --> 19:55.800 to an HTTP API. This costs one order of magnitude less than what we have on the current 19:55.800 --> 20:00.520 resource space. And I think that sometimes, and this is maybe a bit much of a, you know, 20:00.520 --> 20:06.680 I'll comment, but I believe that also in my work, we work with highly potent machines with 20:06.680 --> 20:12.120 lots of resources, but then when we go to the real world, many of the devices that other 20:12.120 --> 20:18.520 our users are using, don't have as many as much process in power for maybe a website or for 20:18.520 --> 20:28.200 maybe anything else, okay. And also, in general, I think that this is a bit of a waste of resources 20:28.200 --> 20:32.200 in the sense that we have a whole Raspberry Pi work we could be using, something that has 20:33.160 --> 20:39.000 a hundred of the computing power and still do the same thing. That's so it's less 20:39.000 --> 20:48.600 egoist, it's a whole of that, right. Then a future work for this would be, it would be very, 20:48.600 --> 20:55.640 very great to be able to create an ISO-115-118 charger that doesn't require a full OS. 20:56.520 --> 21:03.240 And of course, being able to push the wattmeter through mother, because this is like the 21:03.240 --> 21:09.240 new IoT automation standard, and I think there's no definition of a wattmeter in the mother standard. 21:10.760 --> 21:15.480 Please, correct me, I'm wrong if there's anything from the CSA that knows more about this, 21:15.480 --> 21:21.720 but I didn't find a way to publish this information through that, and I believe that for anyone 21:21.720 --> 21:29.160 with a solar, with a solar roof or with any kind of grid management system, it would be great 21:29.160 --> 21:33.480 to be able to automate this in the same way that the rest of the home automation systems work, 21:33.480 --> 21:38.840 and this is like the new standard going forward. And currently, all of the energy-related parts 21:38.840 --> 21:44.520 of balancing are done in a separate protocol, in a separate way, and I think that the 21:44.520 --> 21:50.680 converging could be a good idea, but I'm not sure, and I'm open to questions here, and for 21:50.680 --> 21:57.640 discussion, for the five minutes we have four minutes we have, so yeah, that we could be interesting. 21:59.080 --> 22:06.280 And yeah, in general, just, well, think about all of the CPU hours that we have wasted on 22:08.520 --> 22:13.160 the charges that we have on the streets and the hundreds of euros that every one of them is 22:13.160 --> 22:20.280 costing that we could just remove from existence just by putting something smaller there. 22:21.960 --> 22:26.680 And that's it. Now, if you have any questions, anything to add here, that would be awesome. 22:26.680 --> 22:32.040 Although there's a standard stack which is used for, and when someone has smart leaching or something, 22:32.040 --> 22:36.040 it's not glued to them. I'm thinking to remember what the moment is, there's a standard stack 22:36.040 --> 22:41.160 of including an energy stack, you probably know, which definitely covers metering, because it covers the 22:41.800 --> 22:45.720 metering of a house and so that might be another way of tackling it, because there's a bunch of 22:45.720 --> 22:51.560 software already written around to do some of this sort of thing for you. Okay, so 22:51.560 --> 22:56.040 matter might be too need to be commercially adopted for it, whereas the other stuff is in every 22:56.040 --> 23:02.200 smart metering Europe, for example. What is it called? I'm fading to remember sitting on band 23:02.840 --> 23:08.200 the band 30 or whatever, it's very cool in every consumer access device, so they run this 23:08.360 --> 23:14.040 approach calls. Oh, thank you. Oh, thank you. We might be the last time, 23:14.040 --> 23:24.200 sometimes we can use some of the questions. Yes. All right, so what was being said is that there's 23:24.200 --> 23:30.840 a different energy standard that was older that exists, that is called Zigbee, and in Zigbee 23:30.840 --> 23:38.040 there is an energy profile that you can run and basically, and this is probably a way where 23:38.040 --> 23:43.640 we could go forward with this. I can compliment that with saying that I believe that a lot of this 23:43.640 --> 23:48.600 Zigbee energy profiles are available in matters, so maybe there is a way of doing this. I believe that 23:48.600 --> 23:56.600 this is the case. So yeah, maybe I'm just strong on the success and we can do this. Yeah, yeah, yeah. 23:59.880 --> 24:05.320 Yeah, so anyone has, I don't know if there's anyone from the open EVSE community here, 24:06.200 --> 24:19.240 maybe, no. Okay. Because, yeah, this, I mean, it would be great if we could make this smaller in a way that 24:19.800 --> 24:31.560 is, yeah. Any other questions coming? Yeah, good. Okay, thanks. 24:32.840 --> 24:35.560 All right, thank you.