The problem today: When people are allergic to, for example, gluten, they need to look in the product description of every product in the supermarket to find out if they can eat it or not. That's a problem because it is complicated for users to look at product descriptions of all products.

The solution: To make the food buying experience for users with allergies more simple, we built Filter as a Service, http://www.faasfood.ch/. Filter as a Service helps people quickly find food that accommodates their dietary needs. How did we do it? The filter, we built, filters existing product databases, for example https://www.openfood.ch/. Using the filter, users can simply click on what they want, for example "gluten free products". Gluten free products are then shown immediately.

A little story describing Filter as a Service: Jane goes to the supermarket. She wants to buy frozen french fries and a spices mix product. Jane does not tolerate gluten. To pick frozen french fries and a spices mix product that are gluten free, she reads the product descriptions of the products there. It is complicated for her to find the gluten free products she needs because she needs to read all the product descriptions which are written with small text. On the weekend, a friend of Jane sends her a link to a supermarket that has a Filter as a service solution. Jane's friend writes, "Jane, take a look at this filter solution. When you use the filter, it will be much simpler for you to find gluten free products. Just click on "gluten free products" in the filter, and you will see the gluten free products immediately." Jane tries out the filter and is happy. Her food buying experience just got much simpler. With the filter solution, she can now much better satisfy her dietary needs. And Jane is now even motivated to try out some new gluten free products.

4 photos from the #openfood hackathon: 1: A photo of the filter solution on February 10th: https://twitter.com/ferdinandvogler/status/830115968575733760 2: A photo of the building process on February 10th: https://www.instagram.com/p/BQVpvhclcWC/ 3: A photo of the building process on February 11th: https://www.instagram.com/p/BQXprnrFQOn/ 4: A photo of the Filter as a Service presentation on February 11th. https://www.instagram.com/p/BQX-ckrFICV/

Examples of websites that have served as useful inspiration during the development process: http://www.migros.ch/ http://www.leshop.ch/ http://www.coop.ch/ http://www.coopathome.ch/ http://www.digitec.ch/ http://www.conrad.ch/

Presentation on github: https://github.com/foodandme/foodandme/blob/master/presentation.pptx

food-boarding

Slack-bot to allow people to find a restaurant nearby them that proposes what every diners is eager to eat.

Setup

1. Create a .google file and add in it a google developer key that supports the Google Maps Directions API
2. Create a .slack file and add in it a Slack bot key. Check out the Slack documentation to know how to get one.

How to use

The is just one single command to initiate a lunch you need to ask directly to our bot :

lunch at <time> around <location> with <people> by <transportation>

where : - time is the time of the day (should always be after the current time as we can only use the bot for planning a meal the current day) - location is where you would like to eat : it could be a city, a district, a street etc - people is the people you want to have a lunch with (if many, separate them by a comma); if we detect a channel our bot will notify each of its member individually - transportation is the transportation mode (should be one of driving, walking, bicycling or transit, default to driving)

After that, the person who is starting the lunch and the ones receiving the message will receive some meal photo : each diner will have to share with our bot if he wants to eat that or not, in case he doesn't, our bot will immediately suggest a new meal.

Once every diner has found a meal that fits his desire, our bot inform the table that he founds a restaurant where every one will be able to eat what he wants, and will sends details about the chosen restaurant through a link.

Eventually, a kindly reminder will be send to them 5 minutes before the time they should leave to arrive on time at the restaurant.

How it works

Automated food labels pictures processing involing AI and OCR: -- AI to identify the areas of a label containing textual data (ingredients, nutrition facts, etc; looks like a semantic segmentation problem) -- Textual data area preparation for OCR (rotation, brightness/contrast, etc. processing to remove noise from the text image) -- OCR and text processing to output a structured dataset

The App - What it does

«foodly» is a new application which eases a customer’s life during shopping. We strongly belief that customers won’t scan products just to get additional information about a product without any further benefit. «foodly» combines product transparency by providing valuable information which often isn't easily accessible and by using the data from the scans to simplify and automatize the payment process. «foodly» gives a customer the opportunity to select a product that isn't available at the current store and choose either a store pickup or homedelivery. «foodly» distincts between four types of customers, which have different preferences when it comes to product specification. Customer pick between «Sparfuchs», «Lokalist», «Gesundheit” and «Weltretter». For each type, «foodly» provides relevant information after a scan and helps gaining knowledge in the chosen topic. But it doesn’t stops there, «foodly» also suggest alternative products which fit better with the consumer’s preferences.

The App - How it looks




The App - How you can try it yourself

You can try the App yourself, the APK file for Android devices can be accessed with this link: «foodly» - Android

You can scan one of the following bar codes to test the App yourself:

Blévita 5-Korngebäck


Farmer Crunchy - Honig


M-Classic Orangensaft 25cl


Our Vision

Our vision for purchasing items is that retailers provide applications which supports the whole process from get an idea of what to buy up to the payment and delivery:

(1) In addition to the already clear needs which the customer has, retailers can lead the customer to certain products by providing recepies or additional product suggestions which fit well with already chosen ones.

(2) Most of the time, the same product is offered with different characteristics. Depending on the customer’s preferences, he or she will choose different products and labels. This choice requires different information which often are not quickly recognized or available. The goal of this app is to support the customer in choosing the right product characteristics and optimize his shopping experience. Therefore the customer can, depending on their motivation and goals, choose between four categories: Money, Environment, Health and Local.

During shopping, the customer scans articles mainly to put them into his cart for the self-check-out. In this moment, our app kicks in and provides different types of assistance:
- First, it provides information for the different types, for example it shows the CO2-footprint or the nutrition information of the product. This supports the decision-making of the customer.
- Second, it suggests an alternative product which fits better to the chosen preferences. This automatic alternative substitutes own decision making and doesn't requires attention without losing efficiency in preference matching.
- Third, the app is showing an alarm signal if the customer scans an article which is quite the opposite of his preferences. In a next version an additional feature should be available which checks also if it's against his diet type (i.e. vegan), his religious dietary restriction or his limitations due to allergies.

(3) Customers can choose between picking up their purchases in the shop or use the homedelivery service. Therefore, the scanned products can be ordered via app directly in the online store.

(4) Different Payment methods can be used.

(5) Using the app gives the opportunity to analyze the purchases. Descriptive statistics can show where the customers spend their money or how their CO2-footprint is. In addition, the app can measure how much the alternative product suggestions optimizes the consumers goal.

Our presentation

Our presentation for the hack pitch «foodly» presentation

Who worked on it

Andreas Gassmann, Developer, @AndreasGassmann
Selim Imoberdorf, UI/UX Design - Frontend, @slimiles
Lukas Schönbächler, Developer, @lukeisontheroad
David Schönbächler, Business Development
Patrick Pribilovic, Business Development, @pribilpa
Niklaus Knecht, Visual Design, @niklausknecht
Nicole Strebel, UI/UX Design - Frontend, @enol1c
Samuel Furter, Developer
Pascal Brun, UI/UX Design - Frontend, @IAmPascalB

How to get in contact

If you have any question, you can get in contact with Pascal Brun p.brun@papers.ch

The aim of Foodprint is to raise awareness of the carbon footprint associated with groceries you buy every day and make suggestions on how to reduce the footprint. At the same time, the app helps users to discover new products and recipes. Based on a receipt or collection of purchases, we establish a carbon footprint. The food print can then be compared to previous weeks or other people's footprint. Moreover, the app makes suggestions on how to limit the footprint through smarter purchases such as local or seasonal products.

Intro

Creating a system which allows to track orders in the restaurants. The tracking would be done in real-time and the data would be available freely to everyone (individual orders or aggregated statistics) through a web site/mobile app/api.

To motivate a customer and a restaurant to participate, there is a specific application planned. A customer would be able to see the trending food/orders per restaurant based on number of orders and/or rating of the food.

Application Functionality

• simple management of the current food menu (user: restaurant)

• recording of what have been ordered (user: customer or restaurant)

• rating the ordered food (user: customer)

• providing different statistics about trending orders/food, e.g. currently trending food per restaurant, the best food in the last day, summaries about food categories (user: EVERYONE)

Opportunities (not in scope for Open Food Hackday, save for later)

• trend evaluation through statistics (side-artefact: raw data) as made available freely to the public (e.g. "This Friday night, Zurich has eaten 127 cows, in restaurants)

• application of restaurant to offer rebated prices for meals primarily based on their remaining stock food

• others

Jarvis the Nutritionist

Jarvis is here to help you for your groceries, food questions and so on ! He's smart, kind and neutral, he won't judge you :)

You have two ways of contacting him for the moment. Don't hesitate to drop him a line :

• on Telegram : @JarvisTheNutritionistBot
• on Facebook : Jarvis-le-Nutritionniste

Jarvis understands French and English but will answer you in French.

NB: Jarvis is not online anymore.

Structure

This repo contains some tests about APIs calls and techs. The main thing here is in the api/ folder where you can find the webservices used by the chatbot to interact with the OpenFood API.

The bot itself is in this repo : https://github.com/JarvisMesper/jarvis-the-bot.

Technologies

The chatbot is built with the Microsoft Bot Framework in NodeJS. It uses LUIS (Language Understanding Intelligent Service) to process and understand what the user is asking him. The bot talks to an API made with Flask in Python and hosted on an Ubuntu virtual machine. This API retreive data from OpenFood. All of this is hosted on Azure Bot Service for the demo.

Jarvis is able to understand basic questions like "Who are you ?", both in French and English. He can retrieve product information when the user send him a picture of the barcode using quaggaJS library for the scan.

He also remembers the last scanned products, and let you compare products together.

Context

This project was made during the Open Food Hackdays (10/11 February 2017) oragnized by OpenData.ch @ EPFL.

The following developments of the chatbot are not on the repo. Follow us on Twitter to get notified.

A report containing the state of the project in October 2017 is available here.

Contributors

• Jacky Casas
• Nathan Quinteiro
• Christian Abbet

We described the customer journey of our app in the store and at home.

We pointed out the functionalities of the two databases which are the base of our app: (1) the growing and certified food data base and (2) the growing and open recipe database.

Finally, we identified the key element in order to connect the two databases.

We visualise our project in form of a mobile application. Our goal is to make the origins of Swiss meat more transparent to the end user. In an ideal version, the client should be able to scan a QR code on the meat package, and in return get specific information on the product.

This Info would include: - The farm where the animal has been born (with picture) - It's path of life from birth to shelf - It's name, breed (with picture), height and the age it reached

Our idea is interesting for the consumer, since it makes responsible buying easy. He now has a clear idea of why he might be spending more money for local meat and quality. Another key aspect is that it makes the buying experience personal. While meat is not the product with the lowest CO2 emission, it sure does profit the environment if it is locally produced. With our app, we aim to push consumers to make environmentally friendly choices. For the producer/farmer the app is favourable, since it justifies the price he is demanding for biologically/locally produced meat.

For a prototype, we are using the data traitafina is publicly offering the consumer. Traitafina is a local food maker that guarantees the consumer that he can trace every product from birth to death. This data can be accessed by scanning a code on the product or typing the earmark in on this website: http://www.beefnet.ch/BeefNet/spg. On the other hand we used the "Tierverkehrsdatenbank" (TVB), to complete the information already found on the previously mentioned website.

In the future, we would hope that larger shops, such as Migros and Coop are implementing more exact labeling techniques, so that their products could also be traced.

Project Participants: -Samuel Bühlmann: hello@flaora.ch -Graham Lancashire: glancashire@gmail.com -Martin Holub: mholub.ethz@gmail.com -Doris Fonseca Lima: doris.fonseca.lima@gmail.com

TEAM

• Jonathan Sobel: Bioinfo, Science
• Gianpaolo Rando: Science
• Xavier Stouder : IT Student, full-stack developer
• Jonny Hofmann : IT Student full-stack developer
• Christophe Penot: Marketing
• Sébastien Martin: IT Student and graphic designer
• Nolan Rigo: IT Student (Web Developer)
• Ronald Rodriguez: Chemist
• Luc Patiny: Chemist, Data viz

What is BeerDeCoded?

At BeerDeCoded we map beer at molecular level to help brewers and beer passionates to make more informed decisions.

We want to create a molecular tree of beers that links beers with similar properties. Akin to a fingerprint, this open information could be used for B2B and B2C projects. For instance, BeerDeCoded will help craft brewers benchmark new recipes against existing products and control their quality. Also, the data could train a deep learning sommelier to help beer passionates to discover new beers.

We brainstormed this idea during a past Open Data Hackathon documentation and TODAY WE IMPLEMENTED IT!

Repository:

https://github.com/BeerDecodedProject

Existing Data

We have access to the following datasets:

• List of 240+ Swiss-German Beers google sheets: %Alcohol, Bitternes, .
• Brewing recipes of 200+ BrewDog beers API, Json.
• Untappd social drinking network (max 100 calls per hours) : API
• Beer rating website : API

Fresh Data at the Hackathon

We collected real samples of 118 Beers, crowdsourced from 20 countries tab-separated-values.

We borrowed from the Criminology Department a portable molecular scanner and at the Hackathon we scanned all the 118 beers creating a brand new open molecular dataset for the beer chemical profiles.

Model: First Defender RMX from Thermo Scientific product page A Raman Spectrometer used by the police to identify chemical substances (i.e. explosives, drugs) using Raman spectroscopy

Luc Patiny (EPFL) worked on a webpage to visualise and compare the spectra. Required: Google Chrome online demo.

Page features:

• There is now the colored spectrum 'covariance' that show in red the parts of the spectrum that change the more from one to another
• It also calculates the similarity matrix (just double click on it if it does not appear). Click on the similarity matrix will show you the 2 spectra for the corresponding pixel.
• There is also the 'Clustering' button that allows to display result of hierarchical clustering.

There are traces of the food additive Potassium nitrite (E249, a preservative). In one beer we also found traces of an herbicide.

The tree of beers

We grouped beers with similar molecular profiles into a “tree”.

This dendrogram put beers with similar profiles into different groups (branches). Give us your feedback.

Typical questions with a molecular answer:

B2C - Molecular Sommelier App

Pick a beer you like and see what else is on the branch. Try one beer. Do you like it? Stay close to your comfort zone. Wander away a few branches for something new. Feeling adventurous? Try something on the polar opposite side of the tree.

B2B - Raw material compliance

Is this yeast, hops, wheat or barley really what my provider says it is. Does it contain GMO? Pesticides?

B2B - Food safety

Does my product contain harmful or undesirable microbial contaminants? Does it contain an allergen? Who introduced the contaminant, my production site or my subcontractor? What is the shelf life of this beer?

B2B - Competitor mapping

Does my competitor use my proprietary yeast strain? How does my home-brew beer compare to benchmarks or known brands?

The Beer Market

There are about 20,000 different beer brands on the global market. And the market for craft beers is experiencing a tremendous expansion. In the USA alone, there were more than 1,800 microbreweries in 2014 (+ 27.8% when compared with the previous year) and there are more than 6,000 in Europe and more than 600 only in Switzerland. To differentiate themselves, microbreweries continuously experiment and create new recipes, contributing to as many as 1M of different beer lots brewed every year. This explosion of variety and complexity means that both brewers and consumers face major challenges.

Business Brainstorm:

Who is consuming what? Beer is second after wine! Segmentation of the consumers? Profile utilisateur (as perspectives) Think Mobile First !

Objectives:

• Be able to consult our Beer database on a smartphone
• Be able to compare the prices and order your favorite beer For inspiration: the Ten rules of simplicity (John Maeda)

Segmentation of the database:

• Fermentation process
• Origin of Beers
• IBU, % alcohol, …

Legal question:

• personal data protection
• copyright of beers stickers

Consumers segmentation:

• Age
• Sex
• Taste
• opinion

Data modeling and database engine

The database engine will be RethinkDB. NoSQL advantages + join functionalities.

The BeerDeCoded Challenge @Open Food Data Hackdays

For the BeerDecoded Challenge we did BREW

• Brainstorm about the potential applications of a Beer knowledge-base based on open access data (B2B, B2C)
• Re-analyze some beer data sets
• Produce FRESH DATA open access data set about beers (recipes, sticker, measurements, price, brewery,... )
• Wet-lab - Digitise the molecular spectra of 100+ beers with a portable spectrometer.
• Elucidate the swiss beer market
• Work on a web interface to browse our beer knowledge base USE CHROME. Online demo

The project aims at building a comprehensive database regrouping information about local food producers in order to promote local consumption and short distribution channels. This is to be done in two phases:

• 1 Consolidates existing databases originating from various sources and start a platform to collect up to date and seasonal information.
• 2 Propose a collaborative platform that allows users to complete data as well as a series of apps that make use of this information.

For the Open Food Hackdays, producers from Canton de Vaud are investigated. As a first step, a platform is created for farmers or food enthusiasts to collect and update the information. The database is linked to a map that allows for easy access. The map can be accessed on our homepage.

OpenFarm

In order to support local food producers and to create new economic opportunities for them, we created [OpenFarm]. [OpenFarm] is an open database of points of production & sales (a farm, a market stand, a field), with up-to-date information about available products. This information doesn’t exist today. Consumers have to spend a lot of time searching for it, for example by calling the farmers directly. Having this information more accessible could promote healthier and more sustainable food habits and build a stronger community. It could improve the visibility & traceability of products. It could also provide a channel to sell less desirable products (smaller or damaged for example) and as such reduce food waste.

Local food products need a specific and crowd sourced database because:

• For fresh products, availability is highly seasonal and very dependant on weather conditions
• The points of sale have complex and variable opening hours
• The points of sale are very diverse: self pick in a field, market stand, …

[OpenFarm] answers the following questions:

Where can I buy the products? Is it in the field, in a shop, in the farmyard, in a market?
What is in season now and what is available today? What is the exact variety? Is it bio?
How much would it cost me? Is is cheaper that the supermarket?
When is it open? Do you need an appointment?
How does it works? Do I pick it myself? Is it pre-packaged? Cleaned? Minimum quantity?

The next steps

Currently, we assembled a list of 200+ points of sale by assembling various lists of producers. The data needs to be enriched to include information about the products sold, and needs to be made available with an API. In the next steps, our database could support the following apps. These could be a good challenge for a hackathon once the database is in place.

• An app that signals you that the strawberries in the field next to you are ripe and that you can pick them yourself. The app then takes care of the financial transaction.
• An app that helps you plan a bike route through local farms which have stands open on a specific date and time and that carry the vegetables you are looking for.
• An app that connects farmers and restaurants and helps cooks source their products locally. The app could take care of the delivery and payment.
• An app that allows consumers to rate the products as well as request products, so that farmers know the demand in advance.
• An app that allows farmers to sell products in advance and to set a (flexible) time and date for pick up. If the farmer knows when the customer will come, he can be more efficient.

Additional opportunities of the database:

• Identify gaps in local food supply (quantity, availability) vs. the demand.
• Identify gaps in local food supply (quantity, availability) vs. a specific location.

A project made with the collaboration of:
Aurélie Barberis
Max Carrel
Stefany Krauth
Rémi Tognet
Valentin Tombez
Thomas Wegener

Open Receipts turns your food receipts into actionable data: calories, allergies and more We're using Python/Django as a backend, a Cloud OCR system and the Open Food Facts/Open Food APIs. We get this working on a mobile using HTML5 features (camera pictures, uploads)

• Keep track of food that you purchased and have stored in your home.
• Receive alerts about expiring foods.
• Instead of throwing away food that you won't consume, share it with other users.
• Discover recipes based on the food you have in your home and is close to expiration.
• Keep track of your nutrient intakes.
• Track your CO2 footprint based on your purchases.
• Receive suggestions to replace certain products with environmental friendlier ones.

Presentation: https://docs.google.com/presentation/d/1aYSmKxOMUP0eQS1ycH-8nuMEHW-KqwJgDbv_44Pn3vE/edit#slide=id.p4

SwipEat - Groceries made easy!

Swipeat will make it easy to plan your meals and shop for groceries.

What is Swipeat ?

An application to help you plan meals according to your preferences, and then create the corresponding shopping list.

• Add lifestyle characteristics
• Select meals of your choice or swipe it away
• Shopping list is automatically created, but editing is always possible
• Chosen meals are saved and can easily be found when you will cook

Advantages for the users

1. Time saving
   • No more need to search for recipes, to aggregate ingredients and to write down the list.
2. Easy way to discover new food and recipes
   • Choose in lifestyle characteristic specific food specialties
3. Easy to plan meals for the week
   • Thanks to lifestyle characteristic and history, Swip'Eat provides the user with a lot of recipes.
   • Select or swipe the recipe away.
4. Easy to manage health constraint
   • Health constraints can be added in your lifestyle/preferences
5. Recipes are quickly available at the moment of cooking
6. Shopping list can be edited before, during or after the shopping
   • As soon as something is physically added on the shopping cart, intuitively remove the ingredient from the list
7. The list is automatically categorizing ingredients within food groups vegetables together, dairy, etc.
8. The amount of ingredients is automatically computed depending on the number of person eating a particular meal.

The typical user of Swipeat

• People who don't have time to plan their meals and write down their shopping list
• People who need some culinary imagination
• People who are curious to discover new culinary specialties
• People who want to make it easy to manage their health constraints
• People who are always connected with their smartphone

Targeted Public

• People who do food shopping and who possesses a smartphone

Opensource database

• Swipeat is using open recipe database allowing to build a huge choice of recipes

Development possibilities

• Geo-localization of shops
• Special offers proposed on a specific product in a specific shop (i.e. personalized coupons)
• Recipe database can grow thanks to people who propose new recipes
• Characteristic on a recipe can be added by all users (that allow for instance to classify a recipe as "sport")
• Thanks to the large amount of data gathered on eating and shopping habits, SwipEat will suggest get better at suggestions
• Nutritionist can classify a recipe as "good" for some type of illness
• Online shopping can be done directly within the app thanks to online shopping partners (ie. LeShop)
• Shops can create recipes in order to push some products to targeted customers.
• Producer can become a provider of recipes and so it allows him to attire specific people in his farm for direct sales or for online sales. (thanks to geo localization only user around receive this information)
• Add ingredients on the shopping list and automatically find a recipe with these ingredients.
• Commercial recipes provider provide their recipe database and can get more visibility.

2000kcal Project

PITCH

Flip through Slides Download PDF

PROBLEM

For most people who want to lose weight you are supposed to do only one thing: Consume less calories than you need. This often means:

1. Eat less than 2000 ckal
2. Still eat the right nutritional values.

Trying to figure those two things out is very time consuming, you will have to track every food you eat with tools like http://myfitnesspal.com. These tools will tell you how much calories a certain food has and how much nutritional value it has. But:

1. Its hard to guess how much you actually ate of it. Unless it was the whole can.
2. Those tools won't tell you WHAT to eat.

So you are on your own, you have to create your own diet plans and try to tediously to stick to them.

IDEA: The 2kcal pack

So what if people sat together (think of community) and build daily 2000 kcal Packs that are build out of the existing products available in common super markets. These Packs can be either be:

1. Delivered to your door.
2. Picked up bundled in your supermarket (not very practical see final thoughts below)
3. With the use of an app the consumer can be guided to pick these up at the store.

The benefits are:

1. Its healthy
2. You can be lazy while dieting
3. It helps you to shop the right things.
4. You will save money!

EXAMPLE of a 2kcal Pack

• Caloric Goals: 2000 kcal
• Recommended Nutritional Values (see e.g. http://www.myfitnesspal.com):
  • 125g Carbs
  • 46g Fat
  • 227g Protein
• Optimally this pack is cheap and easy to prepare. Compare 14.95CHF of the pack vs. 60 CHF (10CHF for breakfast at bakery, 25CHF for lunch in a restaurant, and 25 CHF for dinner in a restaurant).

BREAKFAST

2/6 of a pack of croissants: https://www.openfood.ch/de/products/14153 
Calories: 276 kcal (extracted from API)
Nutritional Values: 30g Carbs, 14g Fat, 6g Protein (extracted from API)
Price for whole Package: 3 CHF (extracted via https://produkte.migros.ch/m-classic-buttergipfel-frischback)
-
Coffe from coffe machine at home. 

LUNCH

Whole Pack of  Vegi Bulgursalat mit Hummus-Dip: https://www.openfood.ch/de/products/13264
Calories: 523 kcal (extracted from API)
Nutritional Values: Carbs 14g , Fat 5g, Proteins: 4g  (extracted from API)
Price for whole Package: 6.40 CHF (extracted via https://produkte.migros.ch/vegi-bulgursalat-mit-hummus-dip)
-
Dessert M-CLASSIC KOKOS/SCHOKOLADE JOGHURT: https://www.openfood.ch/de/products/13541
Calories: 223 kcal
Nutritional Values: 27g Carbs, 10g Fat, 5g Protein
Price for the whole Pacakge (1 Yoghurt) 0.55 CHF (extracted via https://produkte.migros.ch/m-classic-joghurt-kokosschokolade)
- 
Tap water

DINNER

Whole Pack of Migros Lasagne 1kg: https://www.openfood.ch/de/products/15016
Calories: 1280kcal (extracted from API)
Nutritional Values: 120g Carbs, 55g Fat, 72g Protein
Price for whole Package: 5CHF (extracted from https://produkte.migros.ch/m-budget-lasagne-bolognese)
- 
Tap water

SUMMARY

Calories: 276+ 253 + 223 + 1280 = 2032 
Nutritional values: 191g Carbs, 84g Fat, 87g 
Cost: 3 CHF + 6.40 CHF+ 0.55 CHF + 5 CHF = 14.95 CHF

• We could say the caloric goal is met < 2k
• We over-ate on carbs(>125g) and fat (>46g), but did not eat enough protein(<227g).
• We only spend 15CHF this day.
• I would rate this pack only 3/5 stars because those conditions are not met

EVALUATION

• Problems:
  • openfood API is incomplete and does not contain all the necessary nutritional values.
  • openfood API does not contain the price of the products.
  • Unsolved how to link between openfood.ch and produkte.migros.ch

FINAL THOUGHTS

• The 2000kcal packs would of course be "virtual", thus on a website or app, since bundling those projects together is not feasible for a supermarket because of different storage conditions (e.g. a yoghurt needs to be stored in the fridge, but you don't want your croissants to be frozen :))
• The app could of course combine the nutritions and the receipe. (e.g. how to prepare a salad using those 3 ingredients)

One end-to-end application to manage your shopping, cooking and diet in four easy steps.

1. Scan products you have at home, in the shop or upload your shopping data (online shopping lists, copies of receipts)
2. Choose recipes among a list of nutritionally ranked choices,
3. Get help to learn new cooking skills
4. Make your shopping list by finding the best nutritional composition for each product.

The app can be tailored to personalized nutritional needs and preferences and will use foods available in your country.

Get ideas on what to cook to reduce food waste, improve your diet, learn how to cook and simplify your shopping.

Please follow the link below for an illustrated presentation on our project (furher details can be also found in the following text): http://bit.ly/2l7RnFV

Context & problem:

Local food markets including so called farmer's market, fresh markets, wet markets, bush markets etc. exist all over the world and they are pillars for urban-rural sustainable development and health. They are crowded and lively places where locals and foreigners sell and buy affordable, nutritious and healthy fresh foods ranging from vegetables, fruits, grains and tubers to poultry, meat, fish and even live animals depending on the socio-cultural context. Due to their strong cultural heritage, local food markets have become major tourist attractions, with travelers visiting them internationally or locally, and discovering local and exotic products and practices. However, information on these markets is very limited and fragmented, and there is no comprehensive map of distribution of local food markets and inventory of their products for Switzerland and the world.

Our solution:

1. Yuzou is the first open and participatory photo inventory of local food markets and food products of Switzerland and the world

2. Yuzou brings together information on the biodiversity of fresh food products including fruits, vegetables, grains, animals etc. with their geographical origin, costs and official information on their nutritional values and health claims that we have obtained from the Swiss Food Composition Database and EU Register of Nutrition and Health Claims.

3. Yuzou is about you and living a fascinating socio-cultural experiences while discovering local food markets on your own or with our local ambassadors in Switzerland and the world

Open data sources:

We combined and cross-referenced three open data sources: iNaturalist.org, the Swiss Food Composition Database and the EU Register of Nutrition and Health Claims. The first source refers to a preliminary project that we developed in 2016 to collect information on products from local food markets.

Economic model and tourism sector opportunity:

We have checked which are the 2017 Global Tourism Trends and Drivers and, guess what? Food tourism, mobile photography and active and responsible tourism are just at the top! (https://www.trekksoft.com/en/blog/7-travel-trends-for-2017-that-will-drive-the-global-tourism-industry)

Yuzou is about living exciting socio-cultural experiences while discovering local food markets in Switzerland and the world. Our platform will build a network of local markets ambassadors and associated activities, that could be linked with existing platforms such as AirBandB (https://www.airbnb.fr/), WithLocals (https://www.withlocals.com/) etc.

The Yuzou Platform:

Our targeted audience are the millenniums travelers looking for authentic food experience around the world. The Yuzou platform allows them to connect with our ambassadors who will guide them through a local market, offer them valuable informations on the local food biodiversity and offer them a unique travel experience. The platform encourage travelers to share pictures & stories of their adventures.

Streams:

• Endorsement (wwf, sierra club, audubon)
• Sponsors ( team with chefs / fauchons supplier / ig foodies & influencers / sustainable brand
• Ambassadors/cicerone
• Green washing opportunities
• Tailormade experience travel
• Visit the farms/vineyards/orchards
• Cooking classes
• Top markets/top destinations
• Public/NGO/nonprofit/state
• Commissions
• Staycation opportunities
• Green B2B i.e. Seed/plant
• Producers selection for Q Bio Certification

Impact:

Our idea is ambicious and innovative and it addresses a Swiss and global gap in information in the food system. The participatory inventory ultimately generated on local food markets and food products will be open and freely accessible to everyone bringing new opportunities for research around Sustainable Development and Public Health. In addition, it will promote local food markets and subsequently local socio-economic development from farm to table, and education and health among the population.

Our team:

We are a highly interdisciplinary team including experts in computers sciences, designers, medical doctors, vets, economists etc. but now we also need your support to develop a beta version of our platform and test it in Swiss local food markets!

• Rafael Ruiz de Castaneda, Isabelle Bolon, Lester Genevieve, Maria Christina Mallet: Public and Global Health, Environmental Health
• Camille Bossel: Arts and Graphic Designer, Marketing
• Sophie Schmidt, Pierre Alexandre Fonta, Gianrocco Lazzari: Computer sciences, Web Development
• Greg Staehli : Marketing, Economy

Most diet-related diseases such as obesity and diabetes are due to a misbalance between several nutrients, not only a few ones. Therefore, the current OpenFood database is incomplete from a nutritional point of view. Only legally required information for the packaging are included (e.g. mainly energy/kcal, proteins, fats and carbohydrates).

We thus aimed at enriching the database through adding nutrient values of already existing food items (e.g. salt content in Gruyeres cheese). Since these data are external (so less reliable), we will tag them.

We used the generic food items from the Swiss Food Composition Database as references and linked OpenFood foods with the most relevant reference food using an algorithm based on food names and nutrient profiles.

Having access to a personal CUMULUS account we are able to calculate the personal daily income of different nutrients for the members in one family. Based on that evidence we recommend a product that will most likely be new for the customer. With a Tinder-like interface, the user can swipe the product to right or left based on the fact if she likes it or not.

Hopefully this will help people discover new products and improve their diet. Take a look to our demo here: http://rebrand.ly/wisepik