View Static Version
Loading

Using Artificial Intelligence for better city planning How a digital tool developed by UNITAC is helping a city to map informal settlements

In a world that is urbanizing rapidly, millions of people are flooding into informal, unplanned settlements, often located at the urban periphery with little or no access to basic services such as clean water, electricity and sanitation. These informal settlements are home of one-quarter of the world’s urban population, which means that around 1 billion urban dwellers live in settlements that have emerged outside of the state’s control.

Estimates show that Africa’s urban population is expected to triple in the next 50 years. The United Nations Sustainable Development Agenda has therefore set the goal of "ensuring everyone has access to adequate, safe and affordable housing and basic services by 2030".

Located in South Africa, the eThekwini municipality has the largest number of informal settlements of any other municipality in the country. It currently houses over 587 informal settlements encompassing 314,000 households, which comprises over a quarter of the city’s population. Continued urbanization and a dwindling supply of well-located serviced land has resulted in a multitude of vulnerabilities experienced by its population.

To address this issue and strengthen community resilience, the city has embraced an ambitious informal settlement upgrading programme, including basic urban service delivery, housing projects and land management mechanisms.

In order to develop a pipeline of urban upgrading and basic urban service delivery projects the city is in need of access to evidence-based information to support the challenges faced in informal areas. It requires up-to-date information, particularly on the number and location of informal structures, as well as environmental constraints that could prevent provision of urban basic services and affordable housing.

The local government has faced various data challenges with regard to informal settlements for evidence-based decision-making, compromising basic urban service delivery.

The challenge of mapping informal settlements

The rehabilitation and improvement of the conditions in informal settlements is an important global challenge. Several government initiatives and a significant amount of effort and resources have been put into this goal. However, accurately mapping and monitoring the growth of informal settlements is a time consuming process and often involves manual digitization of aerial imagery and on the ground surveys.

To support the eThekwini's efforts to improve its data accessibility and management, the UNITAC Hamburg team started a collaboration with the Human Settlements Unit (HSU) of the eThekwini municipality. The objective was to develop an innovative and scalable approach to help the city improve its land monitoring process, particularly focusing on how to better understand informal settlements in the city.

Currently, informal structures are identified by on the ground land monitors – 15 Land Monitors are servicing 587 informal settlements – and through manual marking of structures on aerial photography. This workflow has greatly limited the city’s capacity to respond to resident needs in an efficient and effective way.

eThekwini Kennedy Road informal settlements, South Africa.

Using Artificial Intelligence to address the challenge

To assist the city in automating their building mapping process, the UNITAC team developed a software that uses Machine Learning, especially Deep Learning to radically accelerate the spatial recognition of these settlements and structures based on aerial imagery. In this way, the city can have up-to-date records of the location and extent of its informal settlements, as well as to keep track of changes in the built-up area or density.

BEAM (Building & Establishment Automated Mapper) is an open source and easy-to-use tool that allows the user to quickly detect and visualise the rooftops of buildings in a specific area by simply uploading aerial images of a given location.

Dashboard of BEAM

Through four simple steps, BEAM provides a prediction of the buildings that are present in a given image, technically speaking, it identifies or marks the pixels that the model considers as rooftops. An example can be seen in the pictures below. Aerial or drone images need to be geo-referenced and with a minimum of 72 dpi of resolution. A manual on how to use BEAM is available for download here.

Output images produced by BEAM

BEAM is currently being tested by the eThekwini Municipal Government who has provided feedback and recommendations for improvement. A series of workshops and exchanges with the local team have been planned in order to share the experiences of using the tool and explore further uses and case studies.

Sharing the knowledge

In addition to the development and testing of the BEAM, the second stage of this project, involves the knowledge and skills transfer for the effective integration of the tool within eThekwini’s existing processes and systems. The objective of this collaboration is to enable the eThekwini Municipal Government to better, and more effectively target their upgrading and basic urban service delivery interventions.

Alongside the different departments of the eThekwini Municipal Government, other use cases for the tool will be explored. BEAM may be useful for tackling a number of further challenges related to informal settlement upgrading for more resilient communities, including access to public space and climate adaptation for the urban poor.

Moreover, an Advisory Committee with other interested partners has been established and will profit from the project learnings and experiences throughout the implementation. This committee could also serve as the starting point for upscaling the tool to other cities in South Africa and globally.

In line with UNITAC’s mission to support cities to achieve and accelerate the localization of the SDGs through digital innovation, this project seeks to achieve more efficient and evidence-based planning processes not only in eThekwini, but in other communities, for upgrading and basic urban service delivery interventions in informal settlements. By integrating the perspectives of local stakeholders through a project advisory group, especially of representatives of civil society groups, such as informal area dwellers, human rights groups and transparency initiatives, a people-centred approach will be pursued.

For more information about BEAM visit our website or contact us at unitac@un.org

NextPrevious