The drinking water supply in many developing countries is inadequate, which results in a high incidence of water-borne diseases such as diarrhoea, dysentery, typhoid and cholera. A water-testing method that is accurate, easily accessible and fast is essential for users to make informed decisions on water use.

Currently recommended methods are typically based on culturing, which is accurate, but slow, costly and often require trained personnel, so is rarely used in the communities. In contrast, fluorescence-based methods are fast and easy to use. Though these methods tend to lack specificity, they can be used as early warning and for diagnosis of systems.

Along with the Nepalese partners, we have developed a water assessment system based on low-cost components and can be used with minimal training. It is compatible with the needs of communities and water suppliers in resource-constrained settings, so they can make informed choices about water use, thereby significantly reducing the incidence of water-borne disease.

The system needs to go through piloting cycles, with direct involvement of all of the stakeholders – likely present in the audience –  to optimize the system further according to the local needs and to determine the best strategies to integrate within the current framework of policy and practice.

Diarrhea-causing parasites like Giardia and Cryptosporidium currently requires expensive microscopic examination by trained experts.

an automated system to detect these parasites was demonstrated using AI-assisted smartphone microscopy.

An improved deep-learning models was developed by collecting an extensive dataset of more than 400,000 images from various sources across Nepal and developing web-based tools for data collection and annotation.

The system enables rapid, scalable, and low-cost diagnostic testing, 

A broader vision is proposed for creating a spatio-temporal tracking system for waterborne pathogens using AI-powered smartphone technology and potentially engaging citizen scientists in the process.

Water withdrawal in Brazil has increased steadily, mainly driven by agriculture. Water stress - which is expected to grow steadily - is more pressing issue in Brazil than the CO2 emission - which is expected to not increase significantly.

Agriculture accounts for ~ 80% of total water consumption

Urban and rural supply accounts for~ 25% of total water consumption

Approximately 80% of urban and rural water supply withdrawal returns to water bodies as sewage, of which only 46% is treated

Brazil has vast swathes of vulnerable communities without access to sanitation and water treatment.

There is a pressing need to develop low-cost and portable water sensors to help these communities to monitor the water quality and assist in its treatment.

Nanophotonics - which allows photonic systems to be “printed” in a mass-fabrication process, like roll-to-roll stamp printing - and therefore can be more cost-effective than the traditional optics - is one example where we are working on. These systems can be used in imaging, (eg. large scale multispectral imaging to monitor environment, drought, water stress), microscopy, sensing of the contaminants and as processors in artificial intelligence.