FoodWatch

Food diagnostics: See beyond food

End-to-end food quality & spoilage measurement using mobile-based vision algorithms empowering local stakeholders to save 75% of wasted food

Use Case

Let’s deploy FoodWatch across the value chain. Using FoodWatch, farmers pick the best quality seeds to boost production and optimize harvest times as per ripening of field sections. They detect crop diseases preemptively, saving valuable produce. The biggest leak in this chain, however, is at the storage level. Here, early detection of insect and pest infestation by enterprise deployments of FoodWatch saves millions of tons of staple food grain from being rendered useless. Using FoodWatch, end-users have accurate estimates of the freshness of the food they buy, knowing precisely by when they need to consume it. This tailors consumption to reduce wastage. Excess food that remains is distributed through a sharing economy through C2C models before spoilage. End-to-end, food is preserved.

Potential

While food monitoring is possible with high tech equipment, FoodWatch uses an everyday cameraphone to monitor quality and prevent wastage. With over 70% of the world enjoying access to cameraphones, FoodWatch can potentially ‘regrow’ 7500 trillion kcals of food across the value chain. By processing visual information of upto 500 kgs of food in a few kBs of data in under 15s, FoodWatch tells you expected spoilage dates, optimal storage conditions, and recommends preservation techniques for grains, fruits and vegetables. It enables farmers to select best seeds to maximize harvest (20% saving), optimizes food consumption patterns (20% saving), predicts infestations (35% saving) and distributes unwanted food through B2C/C2C models matching demand to supply (25% saving) - at no additional cost.

Business Case

Farmers, distributors and consumers install the FoodWatch app available for free. The easiest point of entry is cooperatives, after which word of mouth will lead to virality. We do not want to claim any monetary compensation (as % of savings) for the food we save. Revenue is derived from two sources - enterprise solutions and consumers. Enterprise deployments of FoodWatch are available to distributors which are charged for. We also bundle our enterprise deployments with hardware solutions for additional revenue. The main revenue comes from the consumer. Using an Uber-like model, food that would be wasted by a consumer is sold and delivered hyper-locally to a different consumer who uses the food before spoilage. FoodWatch makes a margin of 15-30% on all such B2C/C2C food redistributions.

Objectives:

  1. Saving 35% of food spoiled by infestation during distribution via easy, cheap & routine monitoring of quality & disease at a granular level
  2. Saving 25% of food wasted by consumers via an Uber-like model for real time matching of supply-demand; taking food from wasters to wanters
  3. Improve on-farm yield by 20% - empowering farmers to select the healthiest seeds for sowing, and early prediction of farm diseases and loss

Team Vigyaan

Kharagpur, India

Our Team

Team Vigyaan at the Global Entrepreneurship Summit! / Published January 10, 2016 by Ankur Agarwal

Team Vigyaan at the Global Entrepreneurship Summit!

https://vigyaanforfood.wordpress.com/2016/01/10/team-vigyaan-at-ges/

Team Vigyaan took a trip to the Global Entrepreneurship Summit to meet fellow entrepreneurs and industry veterans to learn from their experiences.

We were a part of GES because we were one of few ventures that was already at the prototype stage.

Through over 50 discussions, we brainstormed different strategies to take our prototype to a broader market in our bid to solve the global challenge of feeding over 9 billion people. We learnt about different challenges that entrepreneurs had previously faced, in an attempt to not repeat them.

Onwards and upwards!

The science behind FoodWatch / Published January 10, 2016 by Ankur Agarwal

The science behind FoodWatch

https://vigyaanforfood.wordpress.com/2016/01/10/the-science-behind-foodwatch/

By the time the naked eye can spot food spoilage it is too late to alter food consumption patterns, or to initiate contact with people who’d be able to consume this food before it rots. End result? The food ends in the trash can. FoodWatch detects spoilage patters well in advance, and allows you to follow consumption patters designed to minimise wastage. Food that can’t be consumed can then be passed to people who need and want it, saving precious income for end users.

FoodWatch captures an image of the food articles we need to analyse using flash cameras and then applies vigorous image processing to predict spoilage trends in those articles and the present stage of spoilage they are in.

As hyper spectral imaging can provide both spectral and spatial information from the tested samples, it can obtain accurate quantification and constituent variation of the food products, leading to better characterisation and improved quality and safety evaluation results. With the integration of the main features of imaging and spectroscopy, hyper spectral imaging can simultaneously acquire both spatial and spectral information that is critical to the detection of food safety and evaluation of food quality attributes. A typical hyper spectral system consists of a light source, a wavelength dispersion device, and an area detector. The images are acquired over the visible and near-infrared (or infrared) wavelengths to specify the complete wavelength spectrum of a sample at each point in the imaging plane. These images are then combined and form a three dimensional hyper spectral cube, with two dimensions for describing spatial information and the third one for spectral information. In this hypercube, each spectral pixel corresponds to a spectral signature (or spectrum) of the corresponding spatial region, recording the entire measured spectrum of the imaged spatial point. Therefore the measured spectrum indicates the ability of the sample in absorbing or scattering the exciting light, representing the inherent chemical properties of a sample. As a result, the technology provides us with unprecedented detection capabilities, which otherwise cannot be achieved with either imaging or spectroscopy alone. The results we get using these techniques are earth-shattering. This technology enables us to figure out whether our fruits our unripe, mature or fully ripe. We are able to detect any signs of spoilage even in fruits or vegetable which otherwise appear absolutely fresh. All this is made possible just by taking photographs at different wavelengths!

Using simple coloured cellophane strips, we can vary the wavelength of light falling on the object by filtering out specific wavelengths.

All of this complex technology is packaged behind a user-friendly interface that enables people at different levels of the food value chain to reap its benefit and save food. After all, a grain of food saved is equivalent to a grain produced.

FoodWatch – taking the cutting edge of technology to the world’s biggest problems.

Expert Opinion: Double testing our prototype / Published January 10, 2016 by Mohit Agarwal

Expert Opinion: Double testing our prototype

https://www.youtube.com/watch?v=7Wa3mI0ADKo

The entire Team Vigyaan got in touch with Soumyadeep Mukherjee, a research scholar studying Computer Vision. Even though our technology team had done a ton of work on the idea and the prototype, we realized that it was easy for the team to get blindsided, and so we went to someone with expertise in the area to get double validation.

He examined our entire code, algorithms, prototype, and our work-in-progress app, and gave us his blessings and suggestions. While the approval was a booster, the suggestions and improvements that he suggested were extremely helpful, and we implemented all of them over the winters.

A small part of our discussion is in the Youtube link attached

Soumyadeep explains that using Hyper Spectral Image Processing allows taking images of fruits and vegetables at the pixel level. All that is required is to take multiple photographs while varying the wavelength of light falling on the object. Processing these multiple photographs gives extremely granular data about the spoilage and infestations of any type of food.

What this means for us is that we can use the flash light in our cell phones coupled with simple cellophane filters of different colours to filter out specific wavelengths and photograph the object. The additional cost is less then 50 cents, and our app can study these wavelength based images to study the food, and that’s when we can drive real impact.

FoodWatch for the win!

Stakeholder Interview V: Time for veges / Published January 10, 2016 by Ankur Agarwal

Stakeholder Interview V: Time for veges

https://www.youtube.com/watch?v=OYiBuk46Tcg

Ankur and Mohit went out to the vegetable market in Kharagpur. We both used our familiarity in Bengali to speak to the oldest vegetable seller in the market.

The nutrition habits of South and South East Asia are very deficient in greens anyway. We further found that greens spoil the fastest, and make for the greatest wastage at the site of the vegetable seller. They must be consumed immediately because of the lack of cold storage in the developing world.

The impact of FoodWatch on greens is tremendous, given their importance in our diet, and the ease of detection of spoilage and infestation in greens due to their monochromic nature.

By saving greens in time, FoodWatch keeps our diets and our bodies, healthy!

Stakeholder Interview IV: A juice vendor shares his point of view

https://www.youtube.com/watch?v=b_x0LBd5Prs

Manav talks to Mr. Ganesh discussing the agents of food spoilage causing the maximum damage. He shared with us details about how he puts in a lot of effort in buying the best produce from the markets, but they rot with time rendering more than 20% of his fruits unfit for consumption.

He was excited to find out that our team is working on a mobile app that will help him solve this issue, using just the cellular phone in his pocket. He wished us success in our endeavours to help solve the global challenge of feeding over 9 billion people.

See beyond food with FoodWatch!

Our Mission

Our team believes that the world’s biggest problems can be solved using technology. We believe that ground level implementation of our technical knowledge across different facets can solve acute problems that plague the world, and more severely, the developing nations. Based out the Indian Institute of Technology (IIT), we understand technology - IIT Kharagpur is one of the world’s leading research institutes, with a dedicated Agriculture and Food Engineering Department and an active Computer Image Processing research centre. More importantly, we understand food and the shortage of it – IIT Kharagpur is a unique amalgamation of pioneering technology located in a town that suffers from acute food shortage. A victim of the infamous Bengal famine of 1943, Kharagpur, in West Bengal, continues to be a poor town with widespread poverty and hunger. Motivated by this marriage, we bring to the table an awareness that only comes from first hand experiences that is missing from urban proponents. We hope to take the sharp edge of technology to solve an everyday problem like food spoilage, wastage and shortage, in turn, making the world a better place.

Our Background

Team Vigyaan is a group of 5 technology and food enthusiasts from the Indian Institute of Technology, Kharagpur pursuing undergraduate and graduate programs across various engineering disciplines. Manav is working towards a Bachelor’s degree with honors in Agriculture and Food Engineering, along with an advanced Master’s degree with his thesis in Food Process Engineering. He brings critical expertise in the visual indicators of spoilage and infestations, having spent his university years on the fields of Kharagpur. Mohit is studying for a Master’s degree in Visual Information Processing and Embedded Systems with his thesis in image correction for aerial pictures. He brings the computer vision knowledge and the ability to deploy software solutions with complex algorithms. Riya is a Bachelor's student, majoring in Computer Science and Engineering. Her expertise at app-development enables the team to convert our technology into mobile-phone applications that are user friendly and can be deployed at all levels of the food chain. Vinit is completing his Bachelor’s and Master’s degree in Economics, and brings to the table a deep understanding of markets, market forces and sharing economies. He is also the modeling brain in the team, allowing us to forecast cause-effect relations. Ankur is majoring in Earth Sciences and lends a crucial element of expertise in soil textures and growing patterns. Moreover, he understands rural-urban interactions, such as that in the food value chain. Living in Kharagpur and seeing first-hand the criticality of the food problem, we want to use our resources to solve the problem of food shortage and wastage. Through TFF, we hope to bring our solution to an international platform. With the critique of the experienced stakeholders and adjudicators at TFF, we will be able to polish our solution further, making it ready for the global markets. Napolean once said – “The world suffers a lot. Not because the violence of bad people. But because of the silence of the good people.” We hope to break this silence.

Team Vigyaan at the Global Entrepreneurship Summit! / Published January 10, 2016 by Ankur Agarwal

Team Vigyaan at the Global Entrepreneurship Summit!

https://vigyaanforfood.wordpress.com/2016/01/10/team-vigyaan-at-ges/

Team Vigyaan took a trip to the Global Entrepreneurship Summit to meet fellow entrepreneurs and industry veterans to learn from their experiences.

We were a part of GES because we were one of few ventures that was already at the prototype stage.

Through over 50 discussions, we brainstormed different strategies to take our prototype to a broader market in our bid to solve the global challenge of feeding over 9 billion people. We learnt about different challenges that entrepreneurs had previously faced, in an attempt to not repeat them.

Onwards and upwards!

The science behind FoodWatch / Published January 10, 2016 by Ankur Agarwal

The science behind FoodWatch

https://vigyaanforfood.wordpress.com/2016/01/10/the-science-behind-foodwatch/

By the time the naked eye can spot food spoilage it is too late to alter food consumption patterns, or to initiate contact with people who’d be able to consume this food before it rots. End result? The food ends in the trash can. FoodWatch detects spoilage patters well in advance, and allows you to follow consumption patters designed to minimise wastage. Food that can’t be consumed can then be passed to people who need and want it, saving precious income for end users.

FoodWatch captures an image of the food articles we need to analyse using flash cameras and then applies vigorous image processing to predict spoilage trends in those articles and the present stage of spoilage they are in.

As hyper spectral imaging can provide both spectral and spatial information from the tested samples, it can obtain accurate quantification and constituent variation of the food products, leading to better characterisation and improved quality and safety evaluation results. With the integration of the main features of imaging and spectroscopy, hyper spectral imaging can simultaneously acquire both spatial and spectral information that is critical to the detection of food safety and evaluation of food quality attributes. A typical hyper spectral system consists of a light source, a wavelength dispersion device, and an area detector. The images are acquired over the visible and near-infrared (or infrared) wavelengths to specify the complete wavelength spectrum of a sample at each point in the imaging plane. These images are then combined and form a three dimensional hyper spectral cube, with two dimensions for describing spatial information and the third one for spectral information. In this hypercube, each spectral pixel corresponds to a spectral signature (or spectrum) of the corresponding spatial region, recording the entire measured spectrum of the imaged spatial point. Therefore the measured spectrum indicates the ability of the sample in absorbing or scattering the exciting light, representing the inherent chemical properties of a sample. As a result, the technology provides us with unprecedented detection capabilities, which otherwise cannot be achieved with either imaging or spectroscopy alone. The results we get using these techniques are earth-shattering. This technology enables us to figure out whether our fruits our unripe, mature or fully ripe. We are able to detect any signs of spoilage even in fruits or vegetable which otherwise appear absolutely fresh. All this is made possible just by taking photographs at different wavelengths!

Using simple coloured cellophane strips, we can vary the wavelength of light falling on the object by filtering out specific wavelengths.

All of this complex technology is packaged behind a user-friendly interface that enables people at different levels of the food value chain to reap its benefit and save food. After all, a grain of food saved is equivalent to a grain produced.

FoodWatch – taking the cutting edge of technology to the world’s biggest problems.

Expert Opinion: Double testing our prototype / Published January 10, 2016 by Mohit Agarwal

Expert Opinion: Double testing our prototype

https://www.youtube.com/watch?v=7Wa3mI0ADKo

The entire Team Vigyaan got in touch with Soumyadeep Mukherjee, a research scholar studying Computer Vision. Even though our technology team had done a ton of work on the idea and the prototype, we realized that it was easy for the team to get blindsided, and so we went to someone with expertise in the area to get double validation.

He examined our entire code, algorithms, prototype, and our work-in-progress app, and gave us his blessings and suggestions. While the approval was a booster, the suggestions and improvements that he suggested were extremely helpful, and we implemented all of them over the winters.

A small part of our discussion is in the Youtube link attached

Soumyadeep explains that using Hyper Spectral Image Processing allows taking images of fruits and vegetables at the pixel level. All that is required is to take multiple photographs while varying the wavelength of light falling on the object. Processing these multiple photographs gives extremely granular data about the spoilage and infestations of any type of food.

What this means for us is that we can use the flash light in our cell phones coupled with simple cellophane filters of different colours to filter out specific wavelengths and photograph the object. The additional cost is less then 50 cents, and our app can study these wavelength based images to study the food, and that’s when we can drive real impact.

FoodWatch for the win!

Stakeholder Interview V: Time for veges / Published January 10, 2016 by Ankur Agarwal

Stakeholder Interview V: Time for veges

https://www.youtube.com/watch?v=OYiBuk46Tcg

Ankur and Mohit went out to the vegetable market in Kharagpur. We both used our familiarity in Bengali to speak to the oldest vegetable seller in the market.

The nutrition habits of South and South East Asia are very deficient in greens anyway. We further found that greens spoil the fastest, and make for the greatest wastage at the site of the vegetable seller. They must be consumed immediately because of the lack of cold storage in the developing world.

The impact of FoodWatch on greens is tremendous, given their importance in our diet, and the ease of detection of spoilage and infestation in greens due to their monochromic nature.

By saving greens in time, FoodWatch keeps our diets and our bodies, healthy!

Stakeholder Interview IV: A juice vendor shares his point of view

https://www.youtube.com/watch?v=b_x0LBd5Prs

Manav talks to Mr. Ganesh discussing the agents of food spoilage causing the maximum damage. He shared with us details about how he puts in a lot of effort in buying the best produce from the markets, but they rot with time rendering more than 20% of his fruits unfit for consumption.

He was excited to find out that our team is working on a mobile app that will help him solve this issue, using just the cellular phone in his pocket. He wished us success in our endeavours to help solve the global challenge of feeding over 9 billion people.

See beyond food with FoodWatch!

Our Team

Our Mission

Our team believes that the world’s biggest problems can be solved using technology. We believe that ground level implementation of our technical knowledge across different facets can solve ...Read More

Our Background

Team Vigyaan is a group of 5 technology and food enthusiasts from the Indian Institute of Technology, Kharagpur pursuing undergraduate and graduate programs across various engineering disciplines. Manav ...Read More

The information contained here represents student project ideas developed as the result of brainstorming activities during Round 1 of the TFF Challenge. It does not represent any final business plans or commercial products.