Accra, Apr 14, GNA – The African Center for Economic Transformation (ACET) and Omdena have completed Africa’s first ever Artificial Intelligence Challenge to predict future infrastructure needs for the continent.
Mr Rob Floyd, the Director and Senior Advisor at ACET, unveiled the findings of the Artificial Intelligence Challenge during a virtual press engagement.
The Artificial Intelligence Challenge was conducted in December 2020, with more than 40 data scientists from 20 countries contributing to the challenge.
Other project partners include the World Resources Institute, Save the Children, and the United Nations World Food Programme.
Mr Floyd said the infrastructure gap facing Africa was huge, up to about $170 billion per year or probably $1.7 trillion over the next 20 years.
He said it would require an investment of about $10 billion dollars to provide internet connectivity to 1.1 billion people by 2030. The challenges were huge and currently 600 million citizens in Africa did not even have access to electricity, he said.
Mr Floyd said only 39 per cent of people in Africa had access to the internet, with only 18 per cent having access to the internet in their homes.
The goal of the challenge was thus to explore how data from artificial intelligence (AI) could be used to predicts Africa’s future infrastructure needs and to help identify priority needs that would yield the greatest impact if invested in.
Among the data experts at the press engagement were Mr Gijs van den Dool, a Data Scientist and Geographic Information System Specialist from France, and Madam Anna Koroleva, a Computational Linguist and Natural Language Processing Specialist and Mr Solomon T. Mulugheta, Data Scientist and Machine Learning Engineer.
Omdena Challenge used publicly available data – satellite images, socio-economic data, climate and topological data, population and demographic data, Google Trends, Google business data, social media data.
The Scientists sought to model the current situation, past temporal changes in population, infrastructure, etc., and in the next step, to predict future demands of infrastructure.
The Omdena Challenge was a ‘proof of concept’ to determine whether AI and machine learning can help predict future infrastructure needs in Africa.
The challenge resulted in innovative models that can be used as tools to make decisions about infrastructure investment. Examples of tools include models that determine water stress, electricity distribution and distance to services.
The project supports ACET’s work on infrastructure and private sector development African infrastructure needs.
Africa’s annual infrastructure financing gap is estimated at $64-108 billion.
Africa’s track record in moving projects to financial close is poor: 80 per cent of infrastructure projects fail at the feasibility and business-plan stage.
Between 2013 and 2017, the average annual funding for infrastructure development in Africa was $77 billion. 42 per cent of which was funded by government budgets.
As a share of GDP, infrastructure investment in Africa has remained at around 3.5 per cent per year since 2000.
ACET hopes that once these models are improved, the would become useful tools for governments and other investors in infrastructure on the continue to make investment decisions that ensure the greatest impact for their citizens and economies.
ACET is an economic policy institute, supporting Africa’s long-term growth through transformation.
GNA