Artificial Intelligence (AI) in the geospatial arena is a combination of computer programming with geospatial analysis, modelling, and visualisation to extract meaningful information from spatial big data. This spans from spatial data modelling to AI methods such as machine learning and deep learning. It also involves automation of data workflows and predictive modelling in various geospatial use cases such as disaster risk management, public health analysis, crop yield estimation, urban growth modelling among many others as seen in recent years.
While appreciating the diversity AI offers it’s important to distinguish the strength of geospatial technology and AI individually and the growth of the GeoAI industry. GIS is a powerful tool mainly used in four areas : visualisation, data analysis, designing data structures and providing decision making frameworks. On the other hand, AI provides a platform for big data analysis and high-end computing processing that support fast deployment of AI models, reducing the time and effort required for manual feature extraction. Thus, combining the two to form GeoAI enhances the accuracy, efficiency, and speed in computation of various applications, and geovisualisation.
As seen with the constant evolution of AI, we can also observe that uses of geospatial data are vast spreading in many domains such as agriculture, urban planning, transportation development, public health, disaster resilience etc. With the overall goal of improving the quality of life by widening the scope of evidence-based decision making and scientific research, GeoAI can be integrated into approaches and workflows for continuous problem-solving.
It offers a number of advantages:
Seamless approaches and new opportunities to integrate and improve current and emergent use of AI technology in geospatial data use cases.
Increased potential in availability of diverse types of geospatial data for different use cases and equally high processing power which is suitable for big data analysis.
A platform for paradigm shift and acceleration in the use of digital transformation tools to address global challenges satisfying the need to have and involve expertise in multiple disciplines to establish best practices on how to approach environmental challenges.
Nevertheless, there still exists several gaps and obstacles such as:
Limited experts and diversity especially in gender representation in the field. This a new technology and many practitioners lack the understanding of the use, software, and skills.
Limited quality checks and verification standards. There is a need for data evaluation and monitoring standards before any analysis to avoid unnecessary costs and bias in data manipulation.
Data privacy and sensitivity. GeoAI technology needs to strengthen the existing privacy measures such as restricting searches of sensitive locations or race discrimination of users and minimise the risk of privacy breaches for new technology.
To move forward in advancing the capabilities that GeoAI technology offers government, private industry players, policymakers and practitioners need to understand the opportunities and risks that GeoAI technology presents and its impact to achieving the United Nations sustainable development goals. Capacity training and knowledge awareness by various academic institutions, NGOs, development agencies, governments, private sectors needs to be enhanced to harness and accelerate collaborative efforts in understanding the impact and growth of GeoAI.
Lastly, we need to ensure that diversity, equity, and inclusion (DE&I) is a fundamental component of geospatial data and AI technology implementation, analysis, resource mobilisation and use. This not only includes diversity in the industry players, but also the involvement of women in training and the decision making process.
About the Author
Stella Chelangat Mutai is a Geospatial consultant for The UN World Food Programme. Stella's work focuses on the Emergency, Monitoring and Evaluation department, covering the regional bureau of Southern and Eastern Africa.