×
Community Blog SaltBae: AI-Powered Salt Production Prediction

SaltBae: AI-Powered Salt Production Prediction

This project is from the team Salt Bae, which was awarded with the First Prize in the Global AI Innovation Challenge 2021 - Intelligent Weather Forecast for Better life.

This project is from the team Salt Bae, which was awarded with the First Prize in the Global AI Innovation Challenge 2021 - Intelligent Weather Forecast for Better life.

Project Introduction

SaltBae is an AI-powered salt production prediction tool for smallholder farmer resilience. By leveraging weather and satellite data, we can empower salt farmers to be more data-driven to increase the sustainability of the business and community.

1
2
3

Solution

1.  Intelligent Sea Surface Salinity Detection

Using data from satellites (Geographical Information System) to measure sea surface salinity and recommend areas with high salinity to salt farmers

2.  Intelligent Salt Production Prediction

Using aggregated data from sea surface salinity detection, weather forecasting and historical salt production level, predict daily salt precipitation rate

3.  Farmer's Decision Support System

Data-driven recommendation to increase salt farmers production capacity E.g: where's the most suitable place to extend the salt ponds?

Technology Highlights

Salt farming is still an overlooked and underserved industry by technology. Most farmers operate with no proper data/information exist to assist their business decision-making. We are combining NOAA GSOD Dataset for the weather data and ESA Sea Surface Salinity Data to answer the question: How can we predict daily salt precipitation rate given the weather, location, and size of farmer's pond using machine learning and give actionable insight to increase their salt-production capacity? And that is SaltBae.

Alibaba Cloud Products Used

1.  Elastic Compute Service (ECS)

Elastic Compute Service was used as an instance to host our application (both front-end and backend).

2.  ApsaraDB

ApsaraDB was used as our relational database to store credentials, user info, and salt production performance.

3.  Domain Name

Domain Name was used to manage www.saltbae.xyz domain name system.

4.  Machine Learning Platform for AI (PAI)

Machine Learning Platform for AI was used to train the salt productivity rate model. We also utilized PAI Studio to frequently update our model

5.  Object Storage Service

Object Storage Service was used as an instance to store our machine learning model for salt productivity rate and credit scoring model

6.  DataWorks

DataWorks was used to orchestrate our machine learning model from training to deployment. Also being used to regularly update our dataset

About the Developer

Hi, we are team Salt Bae from Indonesia; we are both AI enthusiasts with a unique combination of expertise in computer vision, machine learning, and analytical chemistry. Christian is a machine learning engineer with experience in e-commerce, OTA, and fintech companies. Aderian is a pharmacist with experience in a big pharma company in Indonesia. Aderian mostly works on the research, methodology and modeling while Christian works on the product and business analysis.

0 0 0
Share on

Alibaba Cloud Project Hub

113 posts | 24 followers

You may also like

Comments