Leveraging the Cloud for Data and Analytics
CIOREVIEW >> Canada Tech Canada

Leveraging the Cloud for Data and Analytics

Bryan Goodman, Director, AI Advancement Center and Cloud, Ford Motor Company
Bryan Goodman, Director, AI Advancement Center and Cloud, Ford Motor Company

Bryan Goodman, Director, AI Advancement Center and Cloud, Ford Motor Company

My grandfather was a farmer from the 1940s through the 1970s. He witnessed significant technological advancements in farming and farm equipment, and he had to balance his capital investment in equipment to have a successful business. He had to stay modern to be efficient and competitive, but if he spent too much the farm would not have been profitable.

Of course, all companies face similar decisions. When it comes to leveraging the cloud for data and analytics, however, the trade-offs between investment and performance have some interesting characteristics. These make some business decisions rather complex and others quite simple.

At Ford, as in many businesses today, we are well down the path of modernizing all parts of our business, digitally transforming a 118-year-old company into a software-led operation. This transformation is especially important for us to achieve our goal to lead the rollout of new connected vehicle innovations to millions of customers – not just a few luxury buyers – to radically change our customers’ ownership experience — from occasional transactions to an always-on connection with continuously upgraded products and services. In the old way, Ford would make a vehicle, sell it, and hope to see the customer again in a few years when they wanted to purchase their next vehicle. Now, because of vehicle connectivity, we can adapt products to meet customers’ needs, add features and modernize our quality processes through Ford Power-Up over-the-air software updates, and provide services such as predictive maintenance and fleet management.

  Running a modern business without the cloud would be like trying to farm a thousand acres without a tractor 

We must use cloud technologies to deliver these connected products and services. Specifically, we are democratizing AI capability across our enterprise so any member of any team can leverage these powerful tools – whether they are software experts or not. And cloud-based tools have become required for much of today's analytics and AI. Data warehouses — including BigQuery and Snowflake — are good examples. These tools, which are only available in cloud environments, allow Ford experts across our business to query and use large amounts of data without spending months or years engineering and building specialized solutions.

In another Ford example, we have found a key to successful AI projects is to utilize machine learning operations (MLOps). This includes building reliable and repeatable data pipelines, using a modeling workbench to facilitate organization and collaboration, and having robust deployment tools — including monitoring and continuous updating capabilities.

Some of the best tools for data engineering and MLOps run most easily in the cloud, and in some cases they are only available as a cloud service. Apache Beam is excellent for building data pipelines, and Kubeflow is useful for orchestrating machine learning workflows. Tekton works well for cloud-native continuous integration and continuous deployment. For use cases where serverless deployment is an option, the architecture and maintenance are simpler than managing containers or virtual machines.

Operating in the cloud lets us decide where to store data efficiently. This ranges from archival storage for little cost, to high-performance storage with quick processing capabilities.  Running in the cloud also helps us focus on whether compute power for a particular task is worth it, rather than deciding about investing in infrastructure capacity.

The cost of using cloud computing, storage, and services can certainly add up quickly, and we still must optimize and control costs. Many solutions in the cloud offer such improved functionality that the value is well worth it. The total cost of ownership can be lower. A nice benefit is the lack of costly and time-consuming software and infrastructure upgrades. Many services are managed, which further improves efficiencies.

Businesses run on artificial intelligence and business intelligence. They are mission critical. To be done competitively today they require cloud-based tools. Running a modern business without the cloud would be like trying to farm a thousand acres without a tractor.

If my grandfather was farming today, he would be introduced to advanced, connected equipment to help his drive for efficiency by getting the most out of every seed planted and ounce of water used to fuel growth. That is much the same way we are transforming vehicles so customers can use them to make their lives and businesses better, more productive and more enjoyable.

Read Also

The Need For Data Integration And Apis In Today’s Benefits...

Jim Foley, Vice President, Product and Underwriting, Wellfleet Workplace

Data: An Invaluable Asset

Nathaniel Karp, Chief Economist, BBVA Compass

How To Implement A Successful Data Strategy: A Success Story

Elizabeth Puchek, Chief Data Officer,USCIS

Model Implementation: On The Psychology Of Large-Scale Technical...

Thomas Fletcher, PhD, VP Data Analytics, North America Life, PartnerRe

Data Integration Thought Leadership

Matt Meier, EVP, Chief Digital and Data Officer, Driven Brands

Emerging Technologies Driving the Blue Economy

Guillermo Renancio Artal, Director Of Technology, Expansion and Strategic Partnerships, New Pescanova Group