Can AI help to optimize infrastructure planning in Africa?

By Danilo Desiderio

Public infrastructure deficits across Africa are widely documented in economic studies and policy discussions. The African Development Bank’s African Economic Outlook 2018 estimated annual road infrastructure needs at US$130 billion – US$170 billion (2018 USD), with a financing gap of US$68 billion – US$108 billion.

Recently, as of August, the AfDB revised its estimates to US$181 billion – US$221 billion annually. Similar deficiencies exist in critical areas like electricity, internet access, and sanitation.

While these gaps highlight Africa’s development challenges, they also reveal opportunities for high-return infrastructure investments, provided such projects are strategically designed to attract investor interest. Transportation, energy, and water supply projects must undergo thorough evaluations to ensure operational efficiency, cost-effectiveness, and competitiveness, making them more appealing to investors.

Placing infrastructure in strategic locations can further enhance its economic impact.

A World Bank policy research paper from last year emphasized infrastructure’s pivotal role in development. The report noted that public infrastructure, such as broadband internet, increases firm productivity and boosts employment, especially among skilled workers.

Electrification supports structural transformation by enabling shifts from agriculture to higher-value sectors. The paper also noted that railways have lasting developmental effects, as seen in the colonial-era railroads in Africa, which boosted agricultural trade and improved food security.

Countries with more strategically positioned infrastructure tend to experience higher development, better business environments, and stronger logistics performance

Efficiently run ports help reduce geographic isolation, crucial for landlocked and island economies. Roads serve different roles based on their location: rural roads enhance agricultural output and exports by connecting farms to markets, while highways foster manufacturing growth, specialization, and competitiveness.

The World Bank report concluded that recent academic advances, aided by geospatial data, have expanded our understanding of the infrastructure-development relationship. Building on this knowledge, a recent Kiel Institute for the World Economy paper used geospatial data and causal machine learning to explore the economic benefits of targeted infrastructure investments across Africa.

A key finding was that infrastructure in sub-Saharan Africa is often poorly placed, frequently disconnected from economically significant areas. This supports evidence showing that countries with more strategically positioned infrastructure tend to experience higher development, better business environments, and stronger logistics performance. Thus, infrastructure planning in Africa must prioritize spatial precision.

The core challenge is achieving spatial planning efficiency. Machine learning and AI could assist by analyzing complex datasets, including geospatial data, and performing analyses (e.g., cost-benefit, center of gravity, and transportation models) to pinpoint optimal infrastructure locations.

While still in preliminary stages, this approach presents a promising path for future infrastructure planning.

Danilo Desiderio serves as the CEO of Desiderio Consultants Ltd in Nairobi, Kenya, specializing in African customs, trade, and transport policies. He is a customs and trade expert at the World Bank and a senior associate to the Horn Economic and Social Policy Institute (HESPI).