Labels

Insight: AI has the potential to transform and even revolutionise satellite Earth observation in Africa

The combination of AI and satellite data holds great promise for addressing environmental, agricultural and urban planning issues in Africa, making it an essential tool for the development of a continent characterised by its vastness, diversity of ecosystems, varying levels of technological infrastructure and skills. And immense needs. Let us take a brief journey through it.


(c) GAF AG 2018

A vast geographical area

Africa is a vast continent of almost 30 million km2. North America, China and India could all fit comfortably within it. Continuous observation of the continent generates phenomenal amounts of data every day, tens or even hundreds of terabytes. And analysing this data is an even bigger job. A job that will grow mechanically with the arrival of new satellites. 

Range of applications

There are many applications for AI in areas where growing Africa has pressing needs. This is the case, for example, with natural resource management, one of the current areas of intervention for the GMES and Africa programme. 

AI can analyse huge amounts of data much faster than humans and distinguish things that are difficult or time-consuming for us to grasp. It can accurately identify anomalies, unusual variations in vegetation, floods or forest fires. It can detect long-term climate trends, population movements, environmental or urban changes. It can learn to predict future events such as droughts, pest epidemics or environmental degradation.

It can help optimise farming practices by simultaneously analysing data from multiple sources - satellite and in-situ - on soil moisture, crop condition and meteorology, helping to improve yields while reducing costs. It can analyse changes in water levels in rivers and lakes, helping to manage water resources and plan water infrastructure. It can contribute to real-time forest management, quickly identifying areas of illegal deforestation and helping to implement protective measures. It can help identify and protect the habitats of endangered species.

AI is also applicable to disaster management, improving the speed of decision making and minimising the impact. It also has applications in urban planning, at a time when Africa's urbanisation is accelerating and projections show that by 2050 some of today's African capitals will be among the largest cities in the world - Kinshasa, Lagos, Cairo, Dar es Salaam, Luanda - and that this urban expansion will create significant challenges, but also opportunities, in terms of sustainable development, infrastructure and services.

Skills gaps

The continent's vast size and uneven distribution of human resources make it a challenge to effectively monitor and manage its diverse environments and resources. From one country to another (there are 55 countries, all members of the African Union), Africa has sometimes very large disparities in terms of capacity. AI can therefore support and strengthen analytical work to overcome these disparities. It can be used as a transition tool, facilitating training and skills development by making analysis and data accessible to emerging professionals.

Improving access to data

Access to data is another current bottleneck in Africa for a variety of reasons: infrastructure gaps, lack of understanding.  AI platforms can make satellite data more accessible by presenting it in an understandable and usable way. They can enable more users to benefit from satellite data without the need for deep technical knowledge. AI-based systems can tailor data and analysis to the specific needs of African users, be they governments, NGOs or local communities.  

Some examples of current AI applications in Africa

There are many examples of AI applications in Africa. They can be found in personal services and their activities in agriculture (https://hellotractor.com/, http://www.zenvus.com/), education (https://deeplearningindaba.com/), medicine and finance (https://www.jumo.world/). However, Artificial Intelligence (AI) is making significant progress towards application in the field of EO. 

The University of Cape Town is at the forefront of integrating AI into Earth observation. The AI models developed at UCT are particularly useful for tasks such as computer vision, which is crucial for extracting valuable information from satellite imagery. 

The work of the African Institute for Mathematical Sciences (AIMS) includes the development of machine learning algorithms to analyse satellite data.

The African Regional Data Cube, now known as Digital Earth Africa, is working with local partners to use AI to solve problems such as agriculture, deforestation and urbanisation in Ghana, Kenya, Senegal, Sierra Leone and Tanzania - find out more on the Digital Earth Africa website, https://www.digitalearthafrica.org/.

The European contribution, particularly through the European Space Agency (ESA), is also noteworthy. In coordination with the African Union, ESA is supporting several projects and partnerships aimed at harnessing AI for Earth observation in Africa, thereby contributing to the continent's sustainable development goals: Flood forecasting and management, agricultural monitoring, environmental monitoring - see EO Africa R&D research projects, https://www.eoafrica-rd.org/research/research-projects-2023-2024/

Sources















Labels:
Location: Addis-Abeba, Éthiopie

Comments

Post a Comment

Top 10 most viewed articles