It looks like you're having connection troubles. Reconnect now.
01/13/2021 14:49:10
31c3-talk-6158 Latest text of pad 31c3-talk-6158 Saved Jan 13, 2021 |
Hallo Du!
Bevor du loslegst den Talk zu transkribieren, sieh dir bitte noch einmal unseren Style Guide an: https://wiki.c3subtitles.de/de:styleguide. Solltest du Fragen haben, dann kannst du uns gerne direkt fragen oder unter https://webirc.hackint.org/#irc://hackint.org/#subtitles oder https://rocket.events.ccc.de/channel/subtitles erreichen.
Bitte vergiss nicht deinen Fortschritt im Fortschrittsbalken auf der Seite des Talks einzutragen.
Vielen Dank für dein Engagement!
Hey you!
Prior to transcribing, please look at your style guide: https://wiki.c3subtitles.de/en:styleguide. If you have some questions you can either ask us personally or write us at https://webirc.hackint.org/#irc://hackint.org/#subtitles or https://rocket.events.ccc.de/channel/subtitles .
Please don't forget to mark your progress in the progress bar at the talk's website.
Thank you very much for your commitment!
======================================================================
Thank you very much, so I will start my talk by just catching up what I said in the last talked to some who didn't see me yet get get about the project and then I go away from the hardware towards that platform ecosystem because there's a lot of stuff going on in the world right now. And yeah. So basically the idea of the ecosystem is shown in that infographics are made. We have on this and we have the device. This is our current prototype, but there are more and more Internet connected cooking devices coming up. They need some kind of concentrated data, machine readable data, and we want to have the recipes. That's the first coming from the cooks, from cooks all over the world. But we also want to have data about food where we get the ingredients to make that specific recipe and all the available data, nutrition data and data from NGOs. How sustainable is this food? How healthy is this food? And so we can make diet plans and having all this concentrated data, we can also make a great smartphone app independent of having a connected cooking device because we have machine readable data. So, yeah, what's coming first, some about the hardware and then the platform, what's going on in the world and some ideas about how we move on opportunities and needs. So my problem was very simple. I started cooking a result, I went for a quick meal check and it ended up about a bit like this. I was thinking, I need a device that can stir and control the temperature at least. But then starting thinking over the idea was like, no, I don't want a device where I say what it has to do for me, but it's actually the opposite. I want a device that I say, I want to make a risotto. Tell me what you need from me. So in 2009, I was thinking about possibilities. The first CAD model, 2010, I get still virtual prototypes. INSEAD first talks to manufacturers who said it's interesting, but we're still saying it's very futuristic. The first prototype in 2011 looked very dangerous. I remember a startu
p coach telling me you shouldn't have shown me that prototype. I mean, it was a lot of epoxy fiberglass, but I was already cooking under pressure a this on the bottom. This is a load cell. So I had to scale. This is an induction heater I just got from a supermarket with an arduino and a PC at the end. It it worked and it showed that it's technically feasible. We moved on developing prototypes that the fourth generation, it's actually over there at the food hacking base right now, probably making fudge, I'm not sure, but we will make fudge today on this device. So a short overview on the front. We have a power switch. We have function buttons just to to have some direct interaction on device. We do not want to have everything on on a tablet or on a mobile device because you never sure about connectivity. So there must be a way to safely turn the device off. The matrix we had on this device now already has been replaced by a touchscreen. On this version, we have induction, Hillary and of a stainless steel pot we can stir inside the pot. The number 11 is the pusher here it is the pusher. We can push in the vegetables and inside we have a cutting list that tops your onion and yeah, and we have a Raspberry Pi as brain of the device and a Wi-Fi USB stick so we can connect to the world. And we even have a spare Wi-Fi spare USB where we could add Bluetooth, for example, to connect to other connected devices using Bluetooth and of course, Wi-Fi. And so we keep it open to to connect it to all other kinds of devices. Version two of the prototype, we learned we built everything inside a case. It was no more deepsea standing around, but we we still had several issues, but. Then the next version. Could cook for many hours, I mean, last last year, we were cooking, I don't remember how many dishes from tempura to deep fry to to pressure cooking, and we got the whole shield last year and built it into this version. We we used the open source design of the Arduino lol shield to make
that matrix. That's the great thing about having open source hardware you you can use. And we share everything. We made the whole device in welded stainless steel sheet metal, which in the end was not the best idea because through the welding we got lots of issues about tolerances. As soon as you start bringing in heat, the chimney, for example, was not around anymore after welding. So we had to move on from from sheet metal to massive CNCI machined aluminum there. You do not have problems with precision anymore. And as it is, C.A.C. Machining is not really more expensive than welding sheet metal because welding sheet metal is a lot of handwork. S.A.C. machining is is the machine doing the job. And so that's the current prototype. And I think it's pretty close to to release don't see major issues anymore. So we want to to stick with this design and. Yeah, and hopefully bring it out as soon as possible as a developer's kit and later make the HOLSEY certification, we need to be able to sell it fully assembled. So to make it intelligent, of course, we need software inside intelligence means for us, on one hand, minimize the input from the user as much. If the software can anticipate something, we we would like to do it. Of course, there will always be a manual mode where you can override it. But if you do 10 or 100 times the same thing is useful. If the software knows it already, what you want to do and the interactivity, we want to maximize it through, for example, a progress bar for scaling. So you tell us three people are eating. We calculate how many servings, how many of each ingredient you need. And we will show you on the touch screen a progress bar. Add some recent stop. That's enough because we have the scale inside. Uh, timing is also tricky part. So for each cooking step, we assume a cooking time. So you think you know how long it will about take to to cook. And of course, we want to scale this with quantities. So if you want to cut or to peel two kilograms
of carrots, we will have a rough estimate that you won't do it in two minutes. And beyond cooking, shopping, have sustainability. As I said, uh, data about nutrients is easily available. More and more shops have their web shops. But I think it's quite difficult if you have 10 shops which have ten, uh, ten web shops, each one may be changing his available products on a daily basis. And on the other hand, you have a recipe and you have to match that recipe with what's available on the shops. So that's what we want to bring together and of course, sustainability. So you can tell if I'm here in Hamburg and I want to cook a risotto, is there a local farmer who can give me the onions instead of taking onions that come from North Africa or wherever and sharing? Of course, there will be a recording mode so you can try a recipe. If you like it, you upload it to the cloud and all the people can download it, try it and maybe give you some feedback, make it a little bit longer, a little bit shorter, a little bit hotter to to get it better. And I think then through a few iterations we get great recipes and everyone can exactly reproduce it the first time he cooks it. The basic infrastructure we have in the device is currently we have on on the Raspberry Pi, the whole may as well be running because we did not want to reprogram the whole thing we made on the Web. So we just copied the whole Web database onto the device. Of course, later on, we will have to change that. The Raspberry Pi over GPO has a connection to a problem or a microcontroller. I'm evaluating different options for this one. So this one has no electronics inside. It's the case. We working on the electronics because the Raspberry Pi cannot make a real problem. We we need either a microcontroller, a dedicated chip to have a precise figure volume, for example, to control the motor and precise analog to digital converter is also needed specially for the scale. We tried many things, but below 24 beats ADC resolution. T
he scale is quite tricky to make, but there are chips available and we can read them over S.P.I. So we have to load cells, the pressure sensor and the temperature sensor going over that ADC. The interface is now just the Web interface and JavaScript later, we, of course, want to make an API so we can have native apps for better response, responsiveness of of the interface. Um, yeah, we want all data we can get legally and there's a lot of data available. Of course, the recipe is the machine readable recipes. As I said, for each step, we want temperature or pressure. I mean, for pressure cooking. We want the pressure because we can't fix pressure and temperature in the same time or we are over different then of course, the weight of the added ingredients rpm of stirring. How long do you want to stir before we make pause. So we want to make intermittent stirring if possible, and duration of that step and mode modes for the device so the device notes knows what's coming. If we say it's a scaling step, the device goes into scaling mode and for example, turns the heat off to have less interference on the scale. Communication is adjacent, string adjacent. So it's very easy to use it, it's and we build it in a way that we can expand it to zero is the first temperature sensor. P0 is the first pressure sensor, M0 is the first motor. So if you want to build huge devices with many motors and many sensors, you can just expand this. Um, for the ingredients, of course, we also need some data we have what state is it? This is the freshest dry as is frozen. How long can we store it? And we directly link it to nutrient data and later want to link it to products and stores. So if an ingredient, for example, is an onion and we link it to the farmer who sells this onion. Nutrient data, we still use the American data, the Swiss guys now also have released the database. There are nearly each country has a database about nutrient values. Of course, every database comes in a completely dif
ferent format. So it's a bit tricky to get them all together. But we're working on it. But for now, with the US data, we can also already do quite a lot. For the products, as I said, we want to have all products possible, so from the local farmer to the large supermarket, I mean, different people have different wishes, different. And if you have the supermarket in front of the house and have a large family to feed, maybe you just need the the cheap food. And of course, we want to add food that's maybe not accepted by supermarkets. People who have too much food can also feed it into our database and just give it away, because I think it's more useful to to find someone who wants to cook something with it than just to throw it away. We want to. To get all possible feedback, I mean, on one hand, we have veiga on our culture. I call this kind of ethical criteria for food. Then, of course, the the criteria, the how happy were the animals, how how sustainable is it to to eat that that fish, local, organic, gluten free, lactose free. And we want to make a bad system where the users can assign a badge to to some products. And nonprofit organizations also can make that. And so we can quickly react. Maybe if someone is abusing an organic label, also, we we want to react as quickly as possible and through the database, remove that badge if if it was proven that that the food is not organic. The shops, of course, can be from the farm to the supermarket, as I said, each shop, of course, will have GPS coordinates. So it's quite easy to. To find, to shop and to to know to a sign from a recipe, how far do I have to walk to get the ingredients to make the recipe? Now, that's more what I said in the last talk, summed up what's going on in the world. It's it's really interesting to see how how it moved in the last past years. When I started, people were telling me about Jetson's, about future. Crazy. And now more and more people start doing it. Connected devices come on nearly on a mo
nthly basis from connected sweet cookers. I've seen many connected frying pans, connected microwave ovens, connected ovens from many are crowdfunding projects. If you go to Kickstarter and Indiegogo, you see many projects like that coming up. Some are startups, for example, the drop kitchen scale, which is, uh, the Bluetooth kitchen scale, which has a great interface, a great, uh, interactivity. And it's basically made for baking because all that all these baking recipes, you have the flour, the eggs, the milk, and then having a connected scale is very useful. I mean, it tells you, give me flour and just shows your progress, bar stop. And and so you add all ingredients and then you just mix it and and and bake it. But they are limited to that scale. They they think in the future about maybe open it up for other devices I'm in contact with. With the guy from the I have to contact him, but the moment they just launched its scale over all our shops in the US, so he's quite busy, but it's great that that they could start. And we have big brands, big companies also wanting a part of the cake. But still, it's everyone cooking his little own soup, a mess of standards is coming, I think for three devices you will have three recipe platforms and you have three different apps if the manufacturers don't talk to each other. The market leader on the connected on the multifunctional devices here in Europe is forwork, they launched this year a new thermo mix, which now has a touch screen and has recipe chips. So it has some intelligence, but they want to sell the recipe chips together with the cookbooks so they don't tell you what's inside that recipe chips. But it works. I mean, they sell last news. I had to sell the device every 38 seconds. So it works, but it's it's complete. The black box. I contacted them. Do you want to open it up? Do you maybe want to share your recipes with other devices? That's definitely not what they want to do, but I was seeing first posts on forums wh
ere hackers were looking to reverse engineer that recipe chip to look what's inside and how could we add other recipes to that chip? I mean, it's four pins. Maybe it's just USB. I do not have the device, but as soon as I get my hands on, I will try to find out. Then we have from the small devices, we have a grill. It's a grill thermometer, also very, very successful startup company. They are probably more than five years on the market. This is third generation of the Bluetooth thermometer, which is basically a great idea. I mean, you see, you say you have a sirloin steak you put in, you want wanted medium rare. You put in your temperature sensor. You say that on your app and your app calls you when it's time to take your steak off the heat. You have many kinds of meat. But I did not find a way to add and share new kinds of meat, new recipes, and that it's a bit sad. I mean, the guys at the fruit hacking base with with Frantisek, they could use it for fermentation, for example, could make so many things with that temperature sensor if it was not just limited to meat, but now it's limited to meat. They invested quite a lot of money to implement the Apple home kit infrastructure. But I also tried to talk to them. Do you have some kind of recipe database you want to share now that it's not available? Then we have said large manufacturer from France, they launched a few months ago or maybe one month ago that connected to kill the Cuchillo, they made three generations of an intelligent pressure cooker. The first one had 50 built in recipe's. The second one had a USB plug for USB keys to add more recipe, and now the new one has has a Bluetooth connection. They created a project they call Open Food System in 2012 with many partners involved from universities. Specialists around nutrition say it's about more than 10 million euros that were invested, probably also a part coming from from government funding. And they told me it will be finished in 2015, the open food system. B
ut they didn't tell me, will it be buried or will it be launched? Will it be? And it's not so easy to find out what what they mean by open. But what I heard from the project manager of the CUCHILLO is that the app they made for the connected cooker has nothing to do with the open food system. So I was thinking, well, why are you building such a nice ecosystem? And then you make a connected device which with an independent development. And it's amazing to see about the Cuchillo, how communities can develop because there's a huge Facebook community has about 36000 members, uh, French speaking community service, a French company, and they made a huge collection of recipes. That's the index of it. It's a Google doc. If you click off on one of the recipes, you see a jpeg with the how to cook it. And so I think yeah, and it's all in French, of course, so as soon as you don't speak French, you have no access to that recipes. And since this JPEG is just fixed, you cannot scale quantities and so on. So maybe it would be interesting to connect that huge community of users are creating a massive amount of data and and great content to tool called Open Food System or our recipe database or whatever. But it's interesting, they are kind of hacking that device because the original device had 50 recipes. And you see advices like use the broccoli recipe to to make caramel cream or something because it's the same cooking time. So out of these 50 recipes, you can make nearly infinite amount of of new things if you just know, OK, I have to to use the device in that way. Then we have the chef myself community. They also try to make an ecosystem is a consortium also of many companies and other large kitchen device manufacturers. They make gourmet cook. It's actually competitive of the food mix with a nice touch screen with Wi-Fi. This one has Wi-Fi and they now are focused on an ecosystem for older people to to give them give them back some autonomy. But at least with these guys, I am in
in touch and I try to explain them, look, we we should make it only for all the people, we should make it bigger. We should connect all kinds of devices, all kinds of people to that ecosystem to make it more living. And they are thinking about it. The great project, of course, is the open food fact community. Uh, it's a French guy called Stefaan who started that, I think, where you can scan the barcode and you get data about the ingredients. This is mostly focused on prepared foods, not on pure fresh food. But it's a lot a lot of useful data. If you have some kind of sauce and you have allergies, maybe it's useful to know, can I eat this sauce or does it have some ingredients that will cause allergies with me? So, yeah, and it's an it has an open API. So we will try to to use this data with our recipes. Connect this both together for. For helping the user don't Stefanski to continue concerts, Zhenia. And yeah, what? What's all the opportunities, um. We have on the hardware side, if someone wants to make an open source and open source hardware or any device that we could connect to our database because now we have more and more devices, but still with the devices around, I have no idea how to connect them. To, um, to our data, and you'll probably see that more devices, the more people we have, the more it gets. Interesting hacking competitors, of course, I mean, the ones that are close to us, I don't want to them to stay close to us. So I offer one every cook to everyone who opens up a clotheshorse device. So if either it's the Cuchillo or the firebombings or whatever, if if they don't tell us how it works, we will find it out and. Then we have two devices connected to the database and more devices connected to the base, and yeah, we will get a device and if you want, you are on the Hall of Fame of our device of our Web website. If you want to stay anonymous, of course, it's no problem at all on the software side. We also have a lot to do with the text to speech. Ou
tput is somehow working already. Then route planning for shopping and an app for collecting data in the supermarkets because the supermarkets also they are not very willing to to share what is in which supermarket. So it could be the idea to turn it around. Having the camera and the GPS sensor and the barcode reader in each pocket around here, we could get the data and find out what products are available where. And a virtual fridge to know I have this in stock and I should eat it tomorrow because it's going to be bad, it's also a very useful feature we don't have yet. And you probably also have many other ideas how we could grow that ecosystem, make it more useful. Um, yeah, so for this, we would be very happy to find programers, user interface designers. I mean, we have a working prototype, but the user experience is not perfect yet. I'm the electronics mechanics guy and I also do the marketing. I am also the CEO. So it's it's a bit too much. Uh, yeah. Every help is welcome at the moment. Uh, it's difficult to pay salaries if someone of, you know, business angels or investors tell them as soon as we get money, we will pay salaries and. And yeah, as I said on the bottom line, money solves many problems. So that's basically it. Um, we really I'm not sure if the Fajar cooking already, but we will definitely cook them today. So it's at the Fuks hacking base. Just follow this direction along the wall and you will find us. And I also wanted to show you our nice video about the futures, but I was so in my talk that I forgot it. So we can catch this up now. Shouldn't a cooker built for our times do more than just heat up your food? Meet the new every car designed in Switzerland. Every cook includes a powerful heater with precise temperature regulation from melting chocolate at low temperatures to deep frying and high temperatures. You can do everything. We've added a full computer and advanced electronics. We've even left USB ports free to add more features in the future.
The intelligence software connects over Wi-Fi to every recipe store and get you new recipes, you can use your mobile phone to select your favorite ingredients and get the latest suggestions. Every code even gives you a customized shopping list. The powerful motor cut your vegetables directly into the stainless steel pot for cooking. The story keeps moving your food so nothing sticks to the bottom. The story can be intermittent or continuous at different speeds. For Pressure-Cooker, there is a pressure to cover with a safety valve. When the cooking time is over, the program will automatically open the release valve and safely release the steam. Did you know that pressure cooking is two times faster and saves the vitamins? To get the quantities right, we have interactive scaling, every coach shows you a progress bar so you can just pour until it's a stop. Join our community of early adopters. Visitors on every cook dog. And thank you. Thank you. Actually, the video was made with Blendr also opensource and yeah, if you can use Blender, you can make such nice videos. It's a great man in Poland who made it for us. And as I said, it's all open source, all the details you see, they come from the SEAD model and SEAD model is is on grap cards, dot com. It's a great platform where you can navigate through a whole assembly and there you can really watch each screw and open it up, hide some parts and look at some parts. So. Yeah, if if there are mechanical engineers in here, it's always great to have someone looking through and seeing, oh, something to tide here or Toulouse over here. Yeah. So. That's a bit early for the questions, but. I think we can move on to the questions. Yes, please. I think we shall start the Q&A session then right away. Um, one announcement for the audio recording of the whole session. Please do not talk during the question and answer. So the question and answer is for Alexis and the questioners. And because I am standing here, we will start with micro
phone to first, then go to microphone one. And we have something from the Internet on a question from Internet. OK, then please. Microphone to play. Um, nice talk, by the way. And, uh, you mentioned the French recipes and this this Xperia. Have you thought about internationalization of recipe of your recipe database? Yes, that I forgot to mention, but that's a good point. If if you start having a recipe built out of building blocks, you translate the building blocks and you can translate the ten thousands of recipes that are built out of them. And we need to have building blocks to make it machine readable. So currently our database is English and German just to prove it works. But you just add one table for each new language and all recipes get translated. Microphone one, please. I am a chef or I used to be a chef anyway, and my I've got two questions. One is about your device and the expected retail price vs. the art of actually having to cook yourself and learning when to put the potatoes and when to put in the chili and so forth. So that's the first question the economics of your, you know, data centric device versus doing everything in analog format. And the second question is, you specifically targeted the retail market for households. Um, is there a reason why you didn't go for the industrial scale? OK, um. Maybe first to the second question, uh, we we know from we know the size of the retail market, that's the good thing. About three weeks, they publish an annual report and they they sold devices for eight hundred millions last year. So there's certainly a market on the. On the in the restaurants, I am also sure that there is a market, but I, I, I do not have access to it. I think as soon as we have this one here running, I will go to both. I will on one hand put it on the website for for private customers, but I will meet many chefs and talk to them. For a chef, probably five liters is a bit small, but four for a sauce. So five liters also make a lot of ser
vings. So I think for yeah, we could start and figure out how how good is the interactivity? How is it for you as a chef, how how much guidance do you want or how much do you want to be free to, to do what, what, what you like. So what's your estimated retail price. What difficult price? I mean. Um, I would prefer not to say it, because there's still some variables I have quotations ranging from here to here. So if the guy gets it done for the low price, I can make it four thousand five hundred and and still have a little bit of margin left. But if that guy just quoted for too cheap and will deliver poor quality, it will probably be more because it's it's made all of aluminum, it's its strong metal. So it will have some price. But I think for for a professional kitchen where where you have devices going to turn thousands and more, I think it would be a very good price. I've got another question, if nobody has a. Um, microphone four, please. Um, um, I have a first really quick suggestion, but people who you could talk to, I don't know if you ever contacted the food assembly, there might be people who you might be interested in talking to their network of farmers and buyers. And a question, um, about in terms of opportunities, I feel like they could also be, um, a a place for content and information about the foods in terms of, um, patents on seeds and issues such as bio piracy, traditional knowledge of the use of plants. That could be interesting to maybe link through because you're tackling the topic of food. And so you really have a really good access point for this topic. Do you think that could work? Yes, I think that that could be easily integrated into that badge system we want to have. So now it's a question, do you want to have a positive or negative aspect of it? You could make a badge to say it's patent free and GMO free to to have a positive approach to it. But I think that that would be definitely makes sense, because patents on seeds is what I think coul
d also be interesting in that sense would be maybe to actually write the story of the of the ingredients being used and and. Yeah. And the. Yeah. Traditional knowledge, because this disseminating the information also makes it more difficult to patent afterwards because it's been out there in the public domain already now. Yes, we will have to see where we draw the line, I mean, of course, we want to have as much data as possible, but for for some things, we thought we could also just link to Wikipedia and have have just that that existing link. So if you have the data on your platform, maybe linking is a better option than pulling it into our database. But let's stay in touch. That's certainly useful. Thanks. OK, microphone three, please. Hi. So first of all, really big kudos for keeping it open and like working on networking with other projects. That's really great. And my question is somehow around branding, because I see that you work on this hardware product, which is sort of your startup, and this is kind of the commercial aspect in it. And also you work with the knowledge base is data about food and so on. And I wonder how much you would like to associate with your brand and how much you would like to keep more in like neutral this open food facts and open product data that's also Open Food Network, which to some open source software. So the question like, do you have like some kind of interest in keeping this knowledge associated strongly with your brand? Or you kind of feel open to work with more natural brands and like don't focus on your product with branding. We we inside our start up company, we were even discussing about do we make the hardware or do we focus on the platform. So and for the platform, we created the name called Decmil. So digital meal info and we probably will move all the platform from the ever cook domain where it is now to a digital domain to have it independent because it is strongly linked, but it's something completely different. O
f course, one is the Web platform and one is the device. And so the current working title is the Jamil's. We can't find another name for it and we we really want to to keep it open. And yeah. So that's that's the I mean, we we released all the code we wrote until now on the GPL version. Free's probably there's no way back in.
