Garbage sorting in Indonesia: artificial intelligence comes in handy

- Aug 22, 2019-

Since July, when Shanghai introduced "the strictest garbage classification measures in history", garbage classification has become a "compulsory course" in many People's Daily life.Over the past month, the use of new technologies and new models to promote the exploration of garbage classification has been emerging.

The problem of garbage pollution is not unique to China. Many developing countries suffer from it.Google recently revealed in a blog post that an Indonesian startup has used Google's artificial intelligence technology to reduce plastic waste in Indonesia.

Plastic waste is a major problem facing Indonesia.The country has a 50,000-kilometer coastline, and there is a general lack of public awareness about what to do with it, dumping a lot of it into the sea.

Gringgo, an Indonesian start-up, hopes to tackle the problem with the help of technology.

Fibriadi pratama, chief technology officer and co-founder of Gringgo, says the Indonesian landscape makes it difficult to price recyclables.Indonesia contains 17,000 islands, including five major ones, and most of the recycling facilities are located on Java.This increases the cost of transporting recyclables from other islands, so materials with lower recycling values are no longer classified and end up polluting the environment.

Not only that, but sanitation workers in Indonesia often have irregular routes and work schedules, and they lack the knowledge and expertise to identify exactly what items are worth recycling.These factors have a significant negative impact on Indonesia's garbage recycling rate and on the livelihoods of sanitation workers.

In 2017, Gringgo released several pieces of software related to waste management.One of the apps allows sanitation workers to track the type and amount of recycling and recommend more structured routes to save time, allowing manual quantification of recycling and potential revenue.Within a year of launching the software, Gringgo's first pilot village in Indonesia had increased its recycling rate by 35 percent, pratama said.

Earlier this year, Gringgo was named one of 20 recipients of the Google AI impact challenge."We came up with the idea of creating an image recognition tool that could help increase the recycling rate of plastic by sorting different types of waste and determining their value.""Pratama said.

With Google's help, Gringgo is working with another startup to develop image recognition tools using Google's machine learning platform TensorFlow.The goal is for sanitation workers to better analyze and classify waste and quantify its value.

With the blessing of artificial intelligence, sanitation workers will be able to take photos of garbage and determine its value through image recognition."This will educate sanitation workers about the market value of different recyclables and help them optimize their choices and maximize their income.Ultimately, there will be incentives for sanitation workers to recycle and dispose of waste more efficiently and to improve recycling efficiency.""Pratama said.

According to pratama, Gringgo's goal is to continually optimize the ai model to make it more economically sustainable and widely available.

"We believe that ai can help solve the most difficult social and environmental problems of our time -- such as medical care, disaster prediction, environmental protection, agriculture or cultural protection."According to Jeff dean, Google AI principal and senior researcher, Google has launched the "using AI to benefit society" project to explore the use of machine learning related research, which has a positive impact on social, humanitarian and environmental issues.