Manish Jethwa, CTO of Ordnance Survey (OS), shares how the national mapping agency has redefined its digital infrastructure through cloud innovation, AI integration, and a commitment to ethical data practices. From launching the OS National Geographic Database to pioneering machine learning at Cale, OS is setting a new standard for public sector agility and geospatial intelligence
Can you provide an insight into the key milestones in OS’s recent digital transformation journey?
Our digital transformation journey stretches back more than five years. Within this timeframe, Ordnance Survey (OS) took the opportunity to totally redesign our data because customers were asking for faster and easier access. This led to the launch of the OS National Geographic Database (OS NGD), ensuring higher levels of usability, personalisation, and rich attribution. This marked a significant milestone as this was the biggest step change in access to OS data since the launch of OS MasterMap in 2001.
We are fortunate to have our own data centre at Explorer House, complemented by cloud providers such as Microsoft’s Azure. This combination offers flexibility in data storage, ensuring resiliency, high availability, and data bandwidth for processing and updating the OS NGD. By leveraging both on-premises and cloud-based data centres, we can efficiently manage the vast amounts of geospatial data we collect, around 20,000 changes every day.
Today, digital transformation is seen as a continuous improvement cycle that leverages the latest technological advancements to enhance business processes. It’s more about ongoing evolution than a one-time transformation, focusing on staying competitive and adaptable in a rapidly changing technological landscape.
At OS, we are continually investing in both data enhancements and access methods, while some of our data products are released every six months, we are pushing for continuous integration and delivery internally. There is a thirst for accurate and trusted data, which is critical for enhancing public services as it provides the context needed to transform raw data into actionable insights.
OS is one of the most agile government bodies, applying critical technologies that can make a real difference to public services. As part of the trusted digital infrastructure of the country, we translate our changing world into data systems that underpin our economy and communities. By providing our customers with faster and easier access to our data through AI technologies and digital tools, we can streamline processes, support AI innovation, and drive economic growth.
How is OS leveraging AI and machine learning technologies to process and analyse geospatial data?
We use both Computer Vision (CV) and Machine Learning (ML) primarily for data processing, extracting, and analysing features in collected imagery. These advanced technologies enable us to automate the identification and classification of various geographic features, such as buildings, roads, and natural landscapes. By integrating CV and ML into our workflows, we can update the OS NGD more frequently and share changes quickly with customers, ensuring our data remains current. This automation not only improves the efficiency of our data processing but also enhances the accuracy and consistency of the information we provide. As a result, our customers benefit from up-to-date and reliable geospatial data that supports a wide range of applications.
The adoption of AI has increased energy demands for training foundational models in analytics and automation. This trend is likely to continue, but we hope to mitigate it through E F more generalised models supporting various downstream applications. As data centres evolve, they will play a crucial role in enabling real-time analytics, autonomous systems, and other emerging trends in geospatial data. By leveraging advanced data centre technologies, we can enhance our ability to process and analyse large volumes of data quickly and efficiently. The integration of Large Language Models (LLMs) with more elaborate agents will simplify and enhance the way industries approach problem-solving and decision-making.
With the increasing integration of AI in geospatial technologies, what ethical considerations does OS prioritise, and how are these addressed in your AI deployment strategies?
Our commitment to sustainability extends to all aspects of our operations, including data centre management. By implementing energy-efficient technologies and practices, we aim to minimise our carbon footprint and contribute to a more sustainable future. Additionally, we collaborate with our data centre partners to ensure their facilities adhere to high environmental standards. This holistic approach to sustainability helps us meet the growing demand for geospatial data while reducing our impact on the environment.
We consider the ethical implications of data as it moves within our value chain, ensuring that the data is sourced, refined and distributed. To support this, we’ve defined an OS Responsible AI Charter, based on the Locus Charter, to ensure the ethical and responsible development and deployment of AI technologies. By adhering to the principles outlined in the charter, we seek to promote transparency, accountability, and inclusivity in AI practices, ensuring that all stakeholders, especially vulnerable groups, are considered and protected.
We constantly evaluate and adapt this process to maintain the reliability and accuracy our customers expect. By incorporating feedback from users and leveraging advanced technologies like ML, we can continuously improve our validation methods. OS has also created its Responsible AI charter to be mindful of the pitfalls when implementing new techniques, ensuring data privacy, security, and mitigating biases in AI models.
How does OS approach the challenges associated with continuous technological advancements, and what strategies are in place to ensure that both technology infrastructure and workforce adapt effectively?
We are one of the very few organisations that have already deployed CV and ML techniques to explore performing the heavy lifting of feature extraction on images to cover a national scale. Last year was the first time we productionised the workflow to deliver a national dataset to the market and the first time we used ML in that workflow. More of this will happen going forward, to drive currency and more accurate data, and variation in datasets.
The biggest challenges are often encountered at the intersection of data infrastructure and investment in AI. As Generative AI tools continue to develop and proliferate, there is immense pressure to adopt these tools quickly. However, this must be balanced with the need to understand their impact on data management to ensure secure, responsible usage and protection of intellectual property. Overcoming these challenges involves phased implementation, continuous monitoring, and collaboration with cybersecurity and legal experts to maintain a secure and efficient data infrastructure while integrating AI technologies.
We are fortunate within OS that we have an internal Change Management team that ensures our teams are equipped as we progress on our digital transformation journey. To manage the specific implications of AI tooling in any process change, we have included upskilling as a key component of our AI strategy to ensure we manage the impact and risks prior to adoption. Successful transformations rely on a clear vision of future goals, effective communication of progress, and celebrating milestones to sustain momentum.
Can you share examples of how OS collaborates with other organisations in the public sector to drive innovation?
OS collaborates with various public sector organisations to drive innovation through several key initiatives. A notable example is the Public Sector Geospatial Agreement (PSGA), which has significantly improved the use of geospatial data across government and organisations, supporting applications like urban planning, environmental monitoring, and disaster management.
We’ve also adapted our data centre strategy by partnering with leading technology providers, enhancing data storage and processing capabilities. This collaboration further enhances our commitment to sustainability by implementing energy-efficient technologies and working with data centre partners to meet high environmental standards, thereby reducing our carbon footprint.
We support innovation through Geovation, an OS initiative in association with the HM Land Registry (HMLR), that helps geospatial, and property start-ups grow via an Accelerator Programme, Innovation Challenges, and a thriving community. This programme fosters collaboration and drives technological advancements in the geospatial sector.
At OS, we value the importance of cross-collaboration with SMEs, policymakers, and industry partners to drive economic growth and guide effective decision-making based on reliable data and insights. OS is now part of the Department for Science, Innovation and Technology (DSIT) which recognises the importance of geospatial datasets in linking key public sector datasets, unlocking efficiencies, and creating economic value. By embedding addressing data and enhancing customer support, we align with DSIT priorities, aiming to improve citizen experiences and service delivery.
