Could white paint be considered ‘frontier tech’?

Colleagues come together to discuss where the opportunities are to harness technologies to solve challenges across the FCDO. Explore the infographic 📈

Since 2015, the Frontier tech Hub has supported more than 44 pilot projects across 20 countries, all ideated by FCDO colleagues (or as we like to call them, Pioneers). Beyond these ideas are hundreds more that we are unable to take through to pilot stage: 80% of ideas stay on their paper applications and don’t get tested in the real world.

But even on paper, these ideas can show us where the ‘Frontier’ really is.

Some of the pilots in our portfolio use emerging technologies like artificial intelligence, machine learning and big data, but we’re not always looking for these complex technologies. Many of our pilots use mobile money, sensors, nature-based-solutions and other applications that we might not feel are particularly ground-breaking. It’s the environment they’re deployed in, and the nature of the challenge they’re trying to solve that makes them ‘frontier’.

So while the newspapers and big tech editorial websites flash stories about the meta verse, Web 3.0 trends and increasing abilities of machines, we’re far more interested in finding out what opportunities you and your colleagues are spotting across your work.

As an FCDO colleague asked in our Frontier Tech Ideas Showcase yesterday: could white paint be considered ‘frontier tech’?

The answer to this question is yes.

In the Northern hemisphere, a tin of white paint is seen as a cheap, hardware staple, used for decorative purposes. In an equatorial country it has the potential to reflect solar radiation and reduce the need for air-conditioning. Typical commercial white paints get warmer, but innovators have been adjusting their use of the chemical compound barium sulfate to make it more reflective. The result is a paint which reflects up to 98.1% of sunlight, and reduces temperatures at a rate which is comparable with air conditioning units.

“If you were to use this paint to cover a roof area of about 1,000 square feet, we estimate that you could get a cooling power of 10 kilowatts… That’s more powerful than the central air conditioners used by most houses.” Xiulin Ruan, a Purdue professor of mechanical engineering.

So beyond the potential for white paint, what other opportunities are your colleagues spotting across the world?

Despite so many shifts and turbulent priorities occurring across the world during 2022 so far, we received 81 applications to this year’s call for ideas from across 30 countries.

For the first time ever, we invited FCDO colleagues to come and hear what some of those surfacing opportunities are. During the session, 60 FCDO colleagues went into breakout rooms to meet one another, listen and discuss. After 10 minutes we watched them emerge from their virtual spaces, and saw smiles, wide, curious eyes and heard the ends of insightful conversations being cut short as they returned.

Together we explored three of the strongest trends from the applications:

Geospatial modelling to inform environmental management 📡

Around 20% of the applications we received used a range of technologies in order to generate geospatial data models. These technologies included satellites, UAVs and sensors, offering a picture of what is happening in a specific environment at a specific point in time. 

We saw ideas that aimed to generate geospatial data in order to provide decision makers with early indications of natural disasters or risks to key parts of a country’s infrastructure, like drinking water or agriculture. Access to this information would enable the ability to monitor and implement mitigations, and some ideas proposed an additional machine learning component to further make sense of geospatial data and drive things like predictive modelling.

Other ideas sought to quantify the economic value of preserving and sustaining ecosystems through natural capital accounting, and we saw quite a few applications based around the growth of carbon markets. This is an idea we are already exploring within our portfolio, with a pilot using data to support smallholder farmers in Uganda (for whom there are barriers to entering carbon credit markets), to be able to prove their carbon capturing credentials (for things like growing or protecting trees).

A geospatial model which interprets data from sensors on waves, wind, sea levels - as to identify risks of coastal erosion and flooding.

Just under 10% of applications focussed on increasing access to education through E-Learning 🤳🏽

According to UNICEF, during the pandemic at least 463 million children globally were unable to access remote learning after their schools closed. For this reason it’s understandable that we saw an increase in ideas targeting access to education and the digital divide, as well as increasing specific skills such as financial education or STEM for children in schools that lack access to science equipment. 

We saw a number of applications making use of digital platforms, which connect beneficiaries to support or opportunities.

Some were E-learning platforms. These are platforms which are accessed on a device that allows someone to take part in lessons, complete assessments and potentially have some engagement with peers or a teacher. Some applications focussed on E-learning platforms that could be used in classroom settings, with teachers on hand to support students who use them. Data analytics within E-learning platforms bring insights in, covering things like completion rates and progress, which can be used to guide students, or spot students at risk of dropping out or falling behind.

Some applications looked at combining the latest VR technologies to make remote education more compelling. We saw ideas which explored the metaverse, questioning what an immersive, virtual world could do for education. Others looked at hardware options which are more realistic for use in remote schools, asking how distance learning could be improved for communities with technologies which already exist and could work with limited internet connection.

Credit: EdTech Hub

Using AI to improve diagnostics in health 🩺

When we feed machine learning algorithms large amounts of data, it can spot trends and patterns. Those trends and patterns can improve diagnostics in health and save time, money and lives within stretched health systems around the world.

One of our pilots have previously managed to do exactly this. The ‘AI for TB’ pilot explored whether artificial intelligence could be used to accurately diagnose TB and silicosis in South Africa. They implemented a diagnostic triage tool based on a machine-learning model that could detect and assess the presence of TB, silicosis and other respiratory diseases from chest X-rays. 

While the tool was technically successful, the implementing team recognised that it was unable to differentiate TB from silicosis in X-rays to a high degree of sensitivity and specificity. Consequently the pilot concluded that human medical expertise was still required in the diagnosis process. The overall benefit though was that they were able to scan thousands of x-ray images for miners in South Africa - and ensure that those x-rays which showed signs of disease were referred to a doctor for diagnosis - creating a more efficient system, which opened up access to diagnosis (and compensation).

During the process the team explored wider implications and risks of AI in healthcare more generally. What risks are there to privacy, and standards of quality? How do you ensure accountability in health systems when you're potentially looking to outsource decision making to machines?

We want to hold space for more conversations like these, and we know many of you would like to join us too

At the end of the session we asked to know more about the topics you would like to discuss with your colleagues. The chat box filled up with mention of how we support frontier tech advancements, ensuring cyber security is an enabler not a blocker, whether internet connectivity is a critical enabler to frontier tech implementation and whether AI is the future of healthcare.

If you would like to join us then head here to sign up to the Frontier Technologies Network. You’ll receive updates about events and activities available to all FCDO colleagues, to meet and discuss opportunities to harness frontier technologies in your work around the world.

And if you haven’t seen them yet, please do check out the full list of new Frontier Tech Pioneers selected from this year’s applications!