As automotive manufacturers accelerate their shift toward AI, analytics, and software-defined vehicles, a less visible but far more complex challenge is coming into focus. Legacy infrastructure is being pushed far beyond what it was originally designed to handle. Platforms built a decade or more ago, engineered for stability and long lifecycles, are now expected to support GPU-intensive workloads, real-time data processing, and distributed edge architectures.
For Ford Motor Company, this challenge is not theoretical. It plays out daily across a vast and fragmented operational footprint spanning APAC, the Middle East, and Africa. According to Gangadhar Reddy Yasam, Head of IT Infrastructure & Data Centre Engineering and Strategy for APAC and MEA, the hardest part of digital transformation is no longer applications or connectivity, but the foundational platforms that sit underneath them.
“On the infrastructure side, things have evolved quite drastically over the last few years,” says Reddy. While Ford has steadily upgraded networks, endpoints, and Wi-Fi environments, the real constraint lies deeper. The core platforms that run this infrastructure were built many years ago, with the assumption that they would remain largely unchanged for a decade or longer. “You can’t continuously upgrade these platforms every year. They are built once and expected to survive for a long time.”
This creates a structural tension that many large manufacturers are now confronting. AI-led use cases demand modern compute, higher power densities, advanced cooling, and new silicon architectures. But replacing or overhauling core infrastructure at scale, especially in manufacturing environments where uptime is critical, is neither quick nor low risk. As Reddy puts it, infrastructure modernisation in such environments is less about speed and more about sequencing and long-term planning.
The challenge is compounded by geography. Reddy oversees one of Ford’s most diverse regions, covering markets with vastly different levels of infrastructure maturity and partner ecosystems. India, he notes, stands out as a relatively easier environment to operate in. “When I compare India with other countries in the region, the problem here is not so huge,” he says, pointing to the country’s strong OEM presence, partner footprint, and technology talent.
That advantage does not extend uniformly across APAC, the Middle East, and Africa. In several markets, global solution providers lack local presence, forcing reliance on regional or country-specific partnerships. This fragmentation makes it extremely difficult to standardise infrastructure designs or implement a single turnkey model across the region. “We still haven’t been able to crack one turnkey solution partner across the region,” Reddy admits.
As a result, Ford’s infrastructure strategy has had to adapt to local realities. Over-engineered facilities are being right-sized, while older, outdated data centres are being rebuilt to meet current and future demands. At the same time, the company is working with suppliers to help them establish local partnerships so that support models can scale across geographies. Even then, Reddy is candid that this remains an ongoing challenge rather than a solved problem.
Across the global automotive sector, infrastructure strategies are undergoing a fundamental shift as vehicles evolve into software-defined, data-generating platforms. AI-driven use cases, ranging from predictive maintenance and quality analytics to advanced driver-assistance systems, are pushing compute closer to the edge, increasing reliance on distributed data centres. Unlike cloud-native industries, automotive manufacturers must balance these demands against operational continuity, latency constraints, and plant-level autonomy. As a result, hybrid architectures that combine central enterprise data centres with site-level edge infrastructure are becoming the norm, even as they introduce new management and monitoring complexities.
Within Ford, the growing role of AI is inseparable from this infrastructure evolution. Modern vehicles, Reddy explains, are effectively “software-enabled platforms on wheels,” continuously generating massive volumes of data. This data is aggregated into data lakes, but its value depends on the ability to analyse it effectively. “There is definitely a very big need for analytical tools that can take that data and give meaningful insights,” he says, adding that AI is becoming central to unlocking that value.
Autonomous driving is another area where AI is reshaping infrastructure requirements. While fully autonomous vehicles have not scaled at the pace once predicted, many autonomous capabilities are already embedded in today’s vehicles. Reddy believes AI will play a critical role in advancing these features further, even if progress remains incremental rather than revolutionary.
These trends place growing emphasis on edge computing and edge data centres. Many of Ford’s existing edge facilities, however, were built years ago and were never designed for AI-era workloads. “With AI coming in, the consumption requirements of power, cooling, GPUs, silicon are humongous,” points out Reddy. Upgrading edge data centres has therefore become a central focus of Ford’s infrastructure roadmap.
Despite increased adoption of cloud and colocation models across the industry, Ford continues to rely heavily on on-premises infrastructure, particularly for core manufacturing operations. The company follows a hybrid approach, combining enterprise data centres with on-prem facilities at individual plants. Each site operates its own data centre, sized and configured to process locally generated data with minimal latency. This decentralised model supports operational resilience but adds significant complexity from a management perspective.
Looking ahead to 2026, Reddy believes the biggest challenge will not be building new infrastructure, but managing what already exists. “Monitoring one single large, multi-megawatt data centre is one thing,” he says. “But monitoring smaller, fragmented data centres across different countries is a completely different challenge.”
Gaining a consolidated view of performance, efficiency, and health across dozens of distributed edge and on-prem data centres is now a top priority. Managing a single facility is relatively straightforward; orchestrating a geographically dispersed infrastructure landscape is not. For Ford, the next phase of digital transformation will depend less on headline technologies and more on solving this operational complexity—quietly, incrementally, and at scale.