Big Data, Big Benefits for Big Machines

Data analysis lets vehicle operators better understand subtleties of their jobs. (John Deere)

The huge volumes of data created by on-vehicle systems are being mined to bring a range of benefits to vehicle operators and fleet managers. Predictive maintenance is becoming more common while data mining is helping OEM design and manufacturing teams enhance their programs.

Cloud connections let ZF examine huge volumes of data. (ZF)

The seemingly unlimited capabilities of cloud computing let OEMs and fleet managers run big data analysis to understand nuances that previously weren’t even guessed at by experts operating on gut instinct. Data collected from connected vehicles can be scrutinized to spot issues and take care of them.

“We funnel the data from remote diagnostics services back to our engineering group so we can see when issues are occurring and are able to quantify them faster,” said Ash Makki, product marketing manager, connectivity, at Volvo Trucks North America. “With remote diagnostics, we are creating cases which utilize our ASIST platform to include the delivery of repair instructions, parts lists and estimates to the customer.”

To date, the ability to determine when vehicle components or systems are nearing their breaking point is one of the biggest benefits of big data analysis. Fleet operators are using connectivity and the Internet of Things (IoT) to foresee failures.

“Predictive diagnostics eliminates expensive over-maintenances as well as predicts an unplanned breakdown of vehicles,” said Jiri Benes, embedded software engineer at ZF’s digitalization department. “As a result, this proactive approach promises significant cost savings. With our telematics units collecting such information in ZF’s IoT cloud platform based on Microsoft Azure, we are able to analyze large volumes of data for more focused and customizable results.”

A wealth of data

Big data analysis and connectivity are growing together to provide greater insight into many components on the vehicle. Cumulatively, vehicle sensors measure more than a million data points every second, so it’s easy to download a wealth of data from CAN buses.

Data mining techniques make it easy for OEMs to aggregate data from many vehicles to gain new levels of understanding. Similarly, fleet operators can closely monitor their operations and make efficiency improvements.

“Today, the majority of equipment coming from our factories is now equipped with connectivity systems,” said Kevin Socha, connectivity product manager for Cat Digital. “Without this connectivity, the customer and machine are limited to technology within the machine. The data captured comes into a digital platform, which serves as the architecture for running advanced analytics, transforming data and providing common services that are used in our apps. The result is our customers’ ability to gain insight into important information, such as fuel utilization, geolocation or asset operation.”

Big data also helps equipment designers make tweaks that improve operations. Engineers can see how various systems are being used so they can determine how to adjust parameters as diverse as design margins, component lifetimes and user interfaces, to name a few.

“As we develop new products and improve upon existing ones, we use feedback received from vehicles in a number of different ways to create the best equipment possible,” said Doug Sauder, director of applied intelligence, John Deere Intelligent Solutions Group. “While JDLink is beneficial for predictive maintenance, our customers are also using the analysis and information taken from machine data to create the best fleet design for their individual operations.”

Equipment manufacturers are even using big data techniques to improve their own manufacturing operations. Information collected from equipment manufacturing lines and from vehicles can be analyzed separately or in combination to determine how production lines can be enhanced.

Advanced analytics helps Cat equipment operators improve efficiency. (Caterpillar)

“The use of big data analysis in manufacturing can avoid the cost of a reject by either optimizing or eliminating the problematic areas,” Benes said. “Our solution for this is applying so-called digital twinning that represents a virtual replica of a physical asset. Together with an asset tracking system called deTAGtive, we can provide all relevant logistics information as well as track location and condition of these assets in real-time.

Many partners

As with most vehicle technologies today, big data and connectivity have so many aspects that they can only be handled by creating an ecosystem. OEMs are building networks of cloud and app providers, development partners and many other stakeholders. Cooperation with specialists and experts in diverse fields generally brings better results.

“Caterpillar is strategically collaborating with customers, dealers, businesses, start-ups and our own innovation teams to continue our industry leadership by being the disruptor, not the disrupted,” Socha said. “By collaborating with outside businesses, we’re also extending the value of our research and development dollars and maintaining the competitive advantage we need to lead in the digital space.”

Cloud providers offer almost unlimited storage and processing capabilities, making them a central part of these ecosystems. Commercial cloud providers are used by many companies. For example, Microsoft Azure is used by ZF, which runs its own acquisition and analysis programs. Some large OEMs like John Deere handle their own cloud businesses.

In remote areas not currently covered by cellular service, the U.S. government may also be a partner. Companies like Deere advocated for passage of the Precision Agriculture Connectivity Act of 2018, which is now being studied by an FCC commission. That commission’s goal is to conceive a plan for expanding cellular’s coverage to 95% of U.S. croplands by 2025. That will let remote operators send data from the field instead of storing it for periodic uploading to the cloud.