Fraud fighters and bamboo bikes: the African innovators driving change Innovations in development
17 January 2020
Software for fighting cybercrime in Ghana and tools for speeding up cervical cancer diagnosis in Uganda are among innovations recognised by the judges of this year’s Africa prize
The Royal Academy of Engineering’s Africa prize, now in its sixth year, is the continent’s biggest award for engineering innovation. Sixteen African inventors from six countries – including, for the first time, Malawi – have been shortlisted to receive funding, training and mentoring for projects intended to revolutionise sectors ranging from agriculture and banking to women’s health. The winner will be awarded £25,000 and the three runners-up will receive £10,000 each.
This year’s inventions include facial recognition software to prevent financial fraud, a low-cost digital microscope to speed up cervical cancer diagnosis, and two separate innovations made from water hyacinth plants. Four inventors spoke to the Guardian about their innovations and their plans to change Africa for the better.
Ghana: tackling online fraud
Identity fraud and cybercrime are big business in Ghana, where financial institutions spend about $400m (£306m) a year verifying their customers. For Ivory Coast tech entrepreneur Charlette N’Guessan, 25, who led research into what technology Ghanaian banks were using to prevent fraud, the cost was far too high.
“We live in the age of data and fraudsters are getting smarter every day,” says N’Guessan, one of six women shortlisted for the prize. “Online fraud is very high in Africa, and although financial institutions spend a lot of money trying to fix it, they don’t have a real system to prevent it. I thought: ‘I’m a software engineer, let’s talk to banks and see what we can do about this.’”
Along with two friends she met while studying a tech training programme offered by the Accra-based Meltwater Entrepreneurial School of Technology, N’Guessan developed software that uses facial recognition and artificial intelligence to verify identities remotely. Their invention, Bace API, replaces existing methods such as two-step verification or password memorisation. Instead, live images or short videos taken on phone cameras are used to detect whether the image is of a real person, or a photo of an existing image. It then matches the picture or short video to either a pre-saved reference photo, or the person’s government-issued identity documents.
“If it matches, it means the person accessing the services is the same person, it’s not a fake account or a robot,” says N’Guessan. “We have to make sure it’s a human.”
Bace API is currently being used by two local financial organisations. N’Guessan plans to roll out services to more clients within the next three months, primarily in Ghana and Nigeria. She also aims to partner with universities to create a database of students who don’t currently have government-issued ID cards, helping them to gain access to financial services.
“There are so many universities in Ghana, yet so many students miss out on being able to open bank accounts simply because they don’t have passports or driving licences,” she says. “With Bace API they’d be able to use their university ID cards instead. Students would benefit and businesses would tap into a whole new market as well.”
Uganda: cervical cancer screening
In 2016, William Wasswa was preparing medical equipment to help pathologists manually analyse pap smears for cervical cancer. The complicated process was time-consuming and prone to error, and wasn’t helped by long queues of female patients waiting outside the lab for their test results. The experience gave Wasswa, 30, a biomedical sciences assistant lecturer at Uganda’s Mbarara University, an idea how to improve things.
“Cervical cancer is the leading cause of death among female cancer patients in Uganda, and I’ve lost family and friends to the disease,” says Wasswa. “But misdiagnosis is very common. I said to myself: ‘This process could be sped up and improved by using artificial intelligence.’”
Dr William Wasswa
An estimated 100,000 women are diagnosed with cervical cancer every year in sub-Saharan Africa, and 62% are expected to die if they don’t receive treatment. Women who are HIV-positive are five times more likely to develop the preventable disease, which is caused by the human papillomavirus (HPV), the world’s most common sexually transmitted infection. Women in Uganda are disproportionately affected by HIV, accounting for 59% of the 1.3 million adults living with the disease.
Wasswa was inspired to transform the entire process of cervical cancer screening in Uganda. Instead of relying on low-cost microscopes – which, though ubiquitous in labs across the nation, are unable to digitally store samples for analysis – he developed a digital microscope slide scanner to study high-resolution cervical cell images from pap smears, using a 3D printer to make the parts. The device has very good results and only costs a quarter of the price of commercial microscopes, he says.
He then turned his attention to the poor management of patients’ records, none of which are digitised. Again, Wasswa designed specialist software, this time to assess the likelihood of a patient contracting cervical cancer given her risk factors, as well as a separate system for managing and archiving patients’ records using AI.
“Right now, there is almost no follow-up for patients, no reminder when they should come in for their next smear, but our platform also provides telemedicine support,” he says.
The digital microscope slide scanner has performed well and is 75% cheaper than commercial microscopes
Finally, Wasswa created a tool to diagnose and classify cancerous images and plans to establish a repository of images that can be used for cancer research. Overall, his system has had 90-100% accuracy and has been published in peer-reviewed academic journals.
“Our classification system has had very good results, over 98%, but we want to keep improving,” said Wasswa, who is collecting more data to capture the exact workflow of Ugandan hospitals. “We need a dataset based on the Ugandan population, which we can hopefully one day compare with classification models from India or China or the US. And by engaging different stakeholders, such as the ministry of health, we can integrate the system across the whole country.”