To improve public health and sanitation, Thailand’s Ministry of Public Health deployed artificial intelligence (AI) and cognitive services from Microsoft to capture, track and share data on public restrooms through a mobile app.
Leveraging machine learning and computer visualisation, the app helps government agencies identify public health risks and disease hotspots and mitigate the risk of epidemics.
This initiative illustrates how digital transformation can contribute to the social good, and extend the capabilities of government, making a positive impact on the everyday life of its citizens.
Smarter healthcare starts with improved hygiene and public sanitation. By addressing public health risks early on, the public sector can avoid substantial costs later, and save individuals from burdensome medical expenses.
For example, when public restrooms are not kept in optimal condition, they become a breeding ground for diseases, most of which are gastrointestinal, like typhoid fever, a type of enteric fever. Ensuring that public restrooms are clean can significantly reduce or even prevent outbreaks.
Therefore, Thailand’s Ministry of Public Health had a goal to identify aspects of public sanitation that needed immediate improvement.
Microsoft and its partners have been equipping governments and individuals to jointly address public issues with the help of AI technologies, such as machine learning and computer visualisation. Such technologies enable government agencies to get a predictive edge in identifying public health risks and potential disease hotspots.
Advancing through AI
The department of health enlisted approximately 2,000 community volunteers to actively survey the conditions of public restrooms. These volunteers can now utilise a mobile app to capture, track, and share data.
“In the past, public health officers would do surveys and evaluation on paper,” said Siriwan Chandanachulaka, director of the environmental health bureau in the ministry of public health. “Now, they can annotate and import images into the data platform and leverage AI to process the data.”
First, thousands of photos are collected through volunteers’ smartphones and analysed by an image recognition model, which then identifies key feature requirements in toilet hygiene.
According to Oliver Gilbert, managing director at Next Billion, a Singapore-based social enterprise working with the agency, “in just a short initial program, we captured more than 11,000 photos of restroom hygiene in multiple provinces.
“We used these photos to train a sanitation recognition model to identify and evaluate the conditions of public restrooms such as accessibility guard rails, door locks, toilet seats, and flooded floors. And this AI model now delivers about 80–90 per cent accuracy and is improving each day as we train and scale the platform in new environments."
Next, the toilets’ sanitation conditions are scored within the model based on standard government indicators.
Health administrators can identify which aspects of restroom hygiene needs to be improved and notify local restroom operations staff to resolve issues. This data-driven approach can transform public sanitation auditing processes, and public officials can better pinpoint key areas for improvement.
AI is no longer limited to research and development. Instead, these advances make AI more practical and accessible, and put more power into the hands of the community.
Now, the agency can make a positive impact on public health and sanitation, and quickly scale initiatives designed to create a more liveable cities - and improve the lives of citizens.