AI for Optimizing Crop Farming through Sustainable Practices in Africa

AI for Optimizing Crop Farming through Sustainable Practices in Africa

Join a global team of 50 AI changemakers in this high-impact 2-month challenge to optimize crop farming with sustainable practices.

This challenge requires experience in Data Analysis and Machine Learning.

 

The problem

In Sub-Saharan Africa, more than 65% of food produced is from smallholders, but issues such as over-dependence on chemical input are not good for the soil, environment, and food. Soil microbes diminish by continuous use of these chemicals, reduction in biodiversity, and in the long term, the land productivity depreciates such that a higher amount of the chemicals is required to maintain the yield. Small scale farmers do not derive full benefit from their farmlands, as the chemical inputs account for more than 50% of the farm operating cost. These farmers are among the poorest in the society and prices of food keep increasing.  

 

The project goals

The goal is to apply several data analyses and potentially AI-related methodologies to find answers to the following questions: 

  • How to provide ready to apply agro-information for crop farmers (in local language) and farm managers via a dashboard to help them make maximum use of their farmlands
  • How to reduce overhead costs, get better yields, and effectively manage a unique bee-centered cropping system
  • What crops to plant (based on local data from the farms), when to plant, where to plant, when to harvest honey, the quantity of water required to grow the crops, and soil nutrients requirement

 

 

AI for Reducing Food Waste

AI for Reducing Food Waste

A global team of 50 AI change-makers are collaborating in this high-impact 2-month challenge to reduce food waste.

 

The Problem

One-third of all food produced globally is wasted. Contributing to 10% of all greenhouse gases. Reducing food waste is the single greatest action we can take to remove CO2 from the atmosphere.

By giving kitchens full visibility of their food waste, we allow them to see what they’re wasting and how to make changes to their preparation, production, and purchasing habits to reduce it. This will allow companies that serve food to drastically reduce the impact they have on the planet.

 

The Project Goals

We would like to be able to tell the weight of what goes in the bin using just a camera with a scanner.

 

Why join? The uniqueness of Omdena AI Challenges

A collaborative experience you never had in your working life! For the next eight weeks, you will not only build AI solutions to make a real-world impact but also go through an entire data science project lifecycle. This covers problem scoping, data collection, and preparation, as well as modeling for deployment.

And the best part is that you will join a global and collaborative team of changemakers. Omdena AI Challenges are not a competition or hackathon but a real-world project that will take your experience of what is possible through collaboration to a new level.

 

Find more information on how an Omdena project works

 

Using Machine Learning and Drone Datasets to Build a New Segmentation Model for Crop Measurement

Using Machine Learning and Drone Datasets to Build a New Segmentation Model for Crop Measurement

This challenge requires experience in Machine Learning, Image Segmentation, and Data Analysis. The project is hosted with UK-based agtech startup Drone AG.

A team of 50 technology change-makers are collaborating from different experience levels and backgrounds.

 

The problem

This project aims at building a new segmentation/clustering model for the project partner, DroneAg, and their flagship product Skippy Scout, which is a mobile application that automates the flight of a drone to capture data used to provide analysis and insights for farmers and agronomists. Segmentation is used on the captured imagery to measure various aspects of the crop that can be tracked and monitored across multiple flights.

 

The project goals

You will help to build a more accurate segmentation or unsupervised machine learning clustering model that works on a variety of crops, listed below, either by creating a series of models or by building/training a model with a more general training set. We may want to identify and group crops by growth stage rather than crop type. 

The models will be built using a common framework, such as TensorFlow, and can easily be retrained with new imagery (we should also consider a model that actively learns) and can be later iterated upon by our inhouse team. 

The models should be built in a modular way so that we can reuse components for other use cases. One example of this is building each component as a lambda function and orchestrating the overall pipeline with step functions, however, we are open to suggestions for this. 

Further development is likely to involve counting or measuring different parts of the image such as leaves, ears/heads, etc. The models will be accessed by our back-end services via a restful API.

The data and process

Dataset 

You will get access to a number of images of each crop type where we store both the crop type and a plant date (that can help determine the growth stage). These images are unlabeled in terms of segmented values. Also, the results of the current model will be shared, which could be used within an initial training set. 

Crops

Cereals (Wheat, Barley, Oats, etc); OSR; Beans, Peas, Sugarbeet, Soybeans; Corn, Maize; Hemp; Linseed/Flax; Broccoli; Celery; Lettuce; Tobacco; Watercress

 

 

Optimizing Energy Consumption for the Real Estate Market with Computer Vision

Optimizing Energy Consumption for the Real Estate Market with Computer Vision

In this 8-week challenge, 50 AI engineers are collaborating to develop a solution for identifying a building’s different types of elements from 2D images.

This challenge requires experience in Computer Vision, Object Detection, and Machine Learning.

 

The problem

Mechanical/Electrical Design

There is a high cost and long duration to complete a city permit compliant energy model report; it requires at least three months of manual labor for a team of 4-5 people. The current process deters real estate developers from focusing on emission reduction, and the high cost of the model discourages smaller real estate developers from entering the market. This project aims to improve the efficiency of this manual effort by identifying different types of building elements (walls, railings, etc.) in the 2D image.

 

The project goals

This project aims to create an AI system that can accurately identify different types of elements (walls, railings, etc.) in the 2d image.

 

The data

The partner will provide the dataset with the bounding boxes of the relevant type of elements manually drawn.

 

Why join? The uniqueness of Omdena AI Challenges

A collaborative experience you never had in your working life! For the next eight weeks, you will not only build AI solutions to make a real-world impact but also go through an entire data science project lifecycle. This covers problem scoping, data collection, preparation, and modeling for deployment.

And the best part is that you will join a global and collaborative team of changemakers. Omdena AI Challenges are not a competition or hackathon but a real-world project that will take your experience of what is possible through collaboration to a new level.

 

Find more information on how an Omdena project works.

 

Building a Global Company Sustainability Benchmarking System using NLP and Machine Learning

Building a Global Company Sustainability Benchmarking System using NLP and Machine Learning

A global team of 50 AI changemakers are collaborating in this high-impact 2-month challenge to build a sustainability benchmarking system.

This challenge requires experience in NLP and Machine Learning.

 

The context

SustainLab is building an ecosystem of software and AI solutions to help companies become more sustainable. The result of this project, which is related to text mining and text analysis of annual sustainability reports that companies publish globally, will be used to benchmark companies in their industry and globally. The sustainability benchmarking system combined with our software product will be precious for companies. The comparison against competitors is a compelling incentive for companies to set more ambitious goals and take bolder steps towards those goals. As companies are the main contributor to our unsustainable environment, society, and finance, helping them to become more sustainable has a significant impact on our planet and our society.

 

The project goals

A sustainability report is a textual document published by a company about the Environmental, Social, and Governance (ESG) impacts caused by their everyday activities. Most of the time, the companies tend to highlight the positive effects of ESG and may cover up the negative impacts.

This project aims to understand the sustainability work of a company from its textual report and extract valuable data in a structured way that will help in comparing with reports of other companies.

 

Why join? The uniqueness of Omdena AI Challenges

A collaborative experience you never had in your working life! For the next eight weeks, you will not only build AI solutions to make a real-world impact but also go through an entire data science project lifecycle. This covers problem scoping, data collection, and preparation, as well as modeling for deployment.

And the best part is that you will join a global and collaborative team of changemakers. Omdena AI Challenges are not a competition or hackathon but a real-world project that will take your experience of what is possible through collaboration to a new level.

 

Find more information on how an Omdena project works