Driving Responsible AI Innovation: RAIL’s Impact Under AI4D

The Responsible AI Lab (RAIL) at Kwame Nkrumah University of Science and Technology (KNUST) is dedicated to fostering AI research that is ethical, inclusive, and impactful. Launched as part of the first phase of the AI for Development (AI4D) program - AI4D Africa, RAIL has been instrumental in driving initiatives that align AI with local challenges and opportunities in Africa. Through academic training, community engagement, and technological innovations, RAIL has made significant strides in responsible AI development.

This article highlights some of the lab’s key achievements during this phase.

Empowering the Next Generation of AI Researchers

Amina Salifu, a PhD student at Kwame Nkrumah University of Science and Technology (KNUST) and a scholar at the Responsible AI Lab (RAIL), embodies how RAIL nurtures academic excellence and responsible AI innovation.

Amina’s research focuses on developing a Natural Language Processing (NLP) model to detect native and non-native English speakers, with a special emphasis on Ghanaian English accents. By integrating Machine Learning (ML) and deep learning models, her work advances speech recognition technologies tailored to local linguistic contexts, promoting inclusivity and accessibility.

With RAIL’s support—including financial assistance, expert mentorship, and access to research opportunities—Amina has received significant recognition. She won the Poster Presentation Award at both the Deep Learning Indaba in Senegal and the Ghana Data Science Summit (IndabaX Ghana). Her PhD abstract also earned her travel and accommodation waivers to present at the Women in Machine Learning Workshop in Vancouver, Canada, and the African Conference of Women Engineers and Scientists (ACWES) 2024.

RAIL’s initiatives have connected Amina with a global research network, enhancing her academic growth. Reflecting on her experience, she shares:

“RAIL has given my research direction, meaning, and global visibility, equipping me to contribute to responsible AI development.”

Promoting Inclusion Through AI for Persons with Disabilities

In collaboration with the Department of Health Promotion and Disability Studies, RAIL hosted a stakeholder engagement event on 29th November 2024, themed “Promoting Inclusion Through Artificial Intelligence.” This initiative aligns with Sustainable Development Goal (SDG) 10, which seeks to reduce inequalities.

The event convened academia, researchers, persons with disabilities (PWDs), and advocacy groups to explore AI solutions that enhance accessibility. Discussions highlighted the need to design assistive technologies informed by the lived experiences of PWDs, ensuring that AI-driven solutions cater to diverse needs.

Prof. Jerry John Kponyo, Principal Investigator and Scientific Director of RAIL, likened the lab’s mission to a mustard seed, stating:

“Small, purposeful beginnings can transform lives when nurtured through innovation and collaboration.”

The engagement underscored the importance of responsible AI development in fostering a more inclusive society.

Transforming Farming with AI

RAIL’s Crop Disease Detection Toolbox was developed in collaboration with local farmers and the Plant Protection and Regulatory Services Directorate (PPRSD) under Ghana’s Ministry of Food and Agriculture. This innovative AI-powered tool helps farmers identify common crop diseases and provides actionable feedback, mitigation strategies, and treatment recommendations.

To overcome language barriers, the team integrated translation models for Ghanaian dialects, including Asante Twi, Akuapim Twi, Ewe, and Hausa. Additionally, an auto-speech recognition model enables the system to deliver feedback in spoken form, improving accessibility for farmers with low literacy levels.

The Crop Disease Detection Toolbox was further enhanced during the Ghana Crop Disease Detection Challenge, organized in partnership with Zindi, GIZ, Digital Transformation Centre Ghana, and FAIR Forward. Held on 26th November 2024, this event brought together stakeholders from research institutions, government agencies, and the private sector to discuss how AI-driven solutions can mitigate the effects of climate change on agriculture.

Conclusion: The Impact of Multidisciplinary AI Labs

The success of RAIL in its first phase under AI4D underscores the critical role that multidisciplinary AI labs play in driving ethical, inclusive, and impactful AI development. By bridging academia, industry, and government, RAIL has demonstrated how AI can be harnessed to address real-world challenges across multiple sectors—from education and accessibility to agriculture and health.

As AI4D enters its second phase, its commitment to supporting RAIL and similar initiatives remains strong. This next phase will focus on scaling this impact and expanding partnerships across sectors to ensure that responsible AI development continues to thrive in Africa. Through ongoing investment and collaboration, AI4D aims to empower even more researchers, policymakers, and communities with the tools needed to leverage AI for sustainable and inclusive development.