Ford’s relationship with Splunk began in 2013 with connected car dashboards. More than a decade later, that early work has expanded into manufacturing operations, cloud platforms, and large-scale data analysis as Ford pushes AI deeper into how the business runs. Speaking at Splunk .conf25 in Boston, Parshuram Limaye, SVP and Global Head of Cloud Transformation, SRE and Operations at Ford, outlined where that work is focused and how Splunk and Cisco fit into it.
Data at scale
At Ford, data is generated not only by vehicles, but also across people, processes, and manufacturing systems. According to Limaye, the real challenge lies in extracting meaning from that volume.
Limaye said the challenge lies in identifying meaningful correlations across large volumes of data and using those insights to improve customer experience, operational effectiveness, and competitiveness.
Ford is embedding AI across multiple verticals, from manufacturing plants to areas such as digital manufacturing. At the same time, it is addressing how to work with data at scale.
“It’s been a longstanding partnership with Splunk and Cisco. Splunk serves as our data analytics and observability platform for manufacturing plants across the globe, and Cisco, for the most part, provides network bandwidth. We see a lot of data coming from machines, as well as from Cisco devices. We need clear visibility into our manufacturing plants because if there is a stoppage, you need to understand the impact,” Limaye said.
As vehicles move through production, machines are keeping a close watch to keep operations running as expected.
“We can identify, for example, when panels are not fitted correctly or when there is discolouration. Using AI, we are able to detect these issues quickly,” he said.
Unlocking business value
Ford sold 2.2 million vehicles in the United States in 2025, marking a 6% increase in annual US vehicle sales and its highest total since 2019. In the US, the company operates 24 manufacturing plants, including its largest facility in Kansas City, as well as sites such as the Dearborn Truck Plant in Michigan, the Chicago Assembly Plant in Illinois, and the Kentucky Truck Plant in Louisville.
In Asia-Pacific, Ford has at least 12 manufacturing locations, most of them in China, with others in Thailand, Turkey, Vietnam, and India.
“As data comes in, it becomes very important for us to make really smarter and faster decisions. It’s not about managing the data itself, but about understanding what issues the data is showing us. We are also accelerating onboarding for our new commerce initiatives and for customers onto the platform. That is where AI has been particularly useful,” Limaye remarked.
Ford is also evaluating the Time Series Foundation Model, announced last year at .conf25 as part of the Cisco Data Fabric. The architecture is built on the Splunk platform and is designed to process machine data using AI techniques.
Kamal Hathi, SVP and GM at Splunk, said the Time Series Foundation Model is designed for pattern analysis and temporal reasoning on time series data, with applications including anomaly detection, forecasting, and automated root cause analysis across the architecture.
“Time series models at the foundation level are not very common. We started with an open-source model that was tuned more for sensors than for application or network data. That became the basis. We then applied the scale of data needed for this kind of training and tuned the models using sensor and physical data, along with application, network, and security logs,” Hathi said.
He added that the model is intended to help organisations apply AI to their own proprietary data, an area where Ford has significant volumes to work with.
Road to progress
Looking ahead, Ford is exploring how to move intelligence closer to the edge, Limaye said.
“We want a clearer understanding of floor plans, the issues that arise, and how we can process them. We are definitely looking to standardise data, while allowing for local adaptation depending on the region. We are also using SPL2 for data pipeline enrichment. This is part of our ongoing work in this area,” he said.
In closing, Limaye urged organisations to treat data as a shared asset that underpins both efficiency and resilience.
“Don’t treat data in silos. Aim for a shared view across stakeholders, whether that is on platforms, in automation, or within manufacturing plants,” he said.
