Skip to main content
AmigoFish

Openreach drives towards a greener, more connected future with Gemini and BigQuery

From Google Cloud · · 2 min

The player loads from YouTube only when you click. It uses YouTube's privacy-enhanced mode.

If the video does not play, the owner may have turned off embedding. Watch it on YouTube instead.

Watch on YouTube

Not rated yet

Description

*Featured in this video:* Rich Hunt, Data and Insight Director at Openreach

*Executive summary:* Openreach, the UK’s largest wholesale broadband network, partnered with Google Cloud to accelerate its goals of reaching 30 million premises with full fiber by 2030 and becoming a net-zero business by 2031. Using Google Cloud's advanced AI and data technologies, including Gemini and BigQuery, Openreach is applying advanced geoanalytics and empowering planners with a digital replica of the UK to optimize its commercial fleet and enable faster broader installation for millions of customers. These initiatives are already helping to accelerate the rollout of electric vehicles (EVs) — removing roughly 10,000 tonnes of CO₂e off the road every year — and supporting smarter expansion by identifying hundreds of thousands of eligible connections. In addition, Openreach built a custom agent with Agent Development Kit, utilizing Knowledge Catalog as a context engine, to help data engineers convert complex legacy queries into production-ready code for BigQuery, significantly reducing time to insight.

*Challenge:* With a commercial fleet of 20,000 vehicles — the second largest in the UK — driving 200 million miles annually, Openreach aimed to transition from diesel to electric to support its net-zero targets. However, on-premises data infrastructure made it difficult to utilize raw telematics data securely, limiting its ability to map and understand daily driving patterns.

*Solution:* Openreach redesigned its core data foundation on BigQuery, using its enterprise-grade security and default encryption to safely process fleet telematics at scale. With access to global geospatial datasets from Google Maps Platform and Earth Engine directly in BigQuery, Openreach can now layer raw telematics information over zero-emission zones and charging access to optimize routes and accelerate EV adoption. Building on this foundation, Openreach also integrated Gemini to create a digital replica (twin) of the UK, including 35 million premises mapped against national transport infrastructure and existing network assets to pinpoint new coverage opportunities. Critically, this work laid the groundwork for future AI innovation, positioning BigQuery as the single source of truth, with Knowledge Catalog acting as a context engine to automatically aggregate operational knowledge across the organization. Combining this rich contextual layer with the Agent Development Kit (ADK), Openreach deployed a custom AI agent that converts complex legacy queries into clean, production-ready BigQuery code, significantly reducing time to insight for data engineers.

*Results:* With Google Cloud, Openreach is already driving meaningful sustainability and efficiency gains — accelerating its EV transition to remove roughly 10,000 tonnes of CO₂e off the road annually while identifying hundreds of thousands of new full-fiber connections across the UK.

*Key takeaways and highlights from our interview with Rich Hunt, Data and Insight Director at Openreach:*

→ “Raw telematics data of every vehicle is a huge amount of data, we couldn't necessarily give it the right security constraints that it needed to on-prem. We've migrated onto BigQuery where we've started using some of Google's very powerful geoanalytics capabilities. Not only can we just crunch all those numbers really quickly, but we can be confident that we put the right guardrails in place.”

→ “We built a digital replica of the UK within the Gemini platform — every road, railway, waterway network — and we added our network as well. That allows us to do geospatial analysis on how we actually connect customers, and we've managed to add another couple of hundred thousand customers that couldn't access full fiber broadband yesterday that can today.”

→ “Ultimately if you're asking an agent to go and do something, it needs to understand contextually what that data is telling it. Knowledge Catalog, being a contextual brain of the…

Back to Google Cloud · Report a problem