Michael Cavaretta, Ph.D., technical leader of predictive analytics and data mining at Ford Research and Innovation, sees big data as a technology that can solve a lot of internal issues.
Here are some of the highlights from our talk with Cavaretta.Cavaretta (right) led multiple data analytic projects in Ford focused on sales and marketing, warranty and quality, manufacturing, and human resources. Now he's wrapping his head around big data to break down silos inside the company, bolster his data sets by minimizing sampling and finding new insights. The quest: Find Ford's dream data set to predict where the auto market is going and capitalize accordingly.
The border of analytics and big data. Cavaretta said that the last decade at Ford has been spent on analyzing data inside the automaker. "We're just now expanding our vision toward big data," he said. Although big data techniques have historically been tied to melding internal and external information, Ford's focus is largely focused inside the company. In Ford's context, big data would revolve around unstructured data such as log files, diagnostics and mobile sensors, said Cavaretta. The obvious big data return on investment would be warranty costs via proactive fixes.
Storage and big data usage. Cavaretta said the real promise of big data is that a company like Ford can analyze all information instead of relying on samples. "I want to store it all," he said. "Too much data is just lying around. You can now store it and deploy it for benefit." For instance, Cavaretta has been involved in projects where six months of data has been collected and then stakeholders discover that two extra information segments would have been of value. With big data technologies and a pack rat mentality, those wished for data sets could be pulled off the shelf, explained Cavaretta. "Big data is an enabling technology that means you don't have to sample," he said.
The promise of big data. "The most significant idea for big data is that it allows you to see around corners and react," said Cavaretta. For instance, Ford can aggregate customer feedback and internal logs to add new features and technology to its cars.
An analytical culture. There has been a lot of talk about the shortage of data scientists and big data experts, but Cavaretta noted that it may be more important to give the rank-and-file employee information in the field. "Not everyone has to be able to create statistical models, understand them and slice and dice data quickly," said Cavaretta. "We need more people understanding data in general."