Volvo Trucks makes increasingly smart vehicles.
More than 350,000 Volvo rigs traverse North American highways every day, equipped with IoT sensors that monitor conditions and send data for troubleshooting and analysis. Embedded telematics allows over-the-air updates to engine software. The combination of on-board technology and a back-end analytics platform enables Volvo Trucks to process millions of data records instantly. Using IoT and artificial intelligence, Volvo Trucks is able to reduce diagnostic time by 70% and truck repair time by 25%.
For fleet managers, the biggest benefit is improved uptime. “Through more efficient and proactive maintenance, Volvo Trucks can help customers maximize the time their vehicles are on the road and minimize the cost of service interruptions,” said Conal Deedy, director of customer productivity solutions for Volvo Trucks North America, the company’s Manufacturing company Volvo Group, part of Swedish multinational corporation.
Volvo Trucks’ vehicle connectivity features have been developed for eight years. Its first foray into connected trucks began in 2012, when it developed a remote diagnostic service that allowed customers to transfer information between trucks and monitor mechanical fault codes in real time. Shortly after launching the diagnostic service, Volvo Trucks launched a remote programming service that allows over-the-air software and parameter updates via cellular services.
A fault code is triggered when there is a problem with one of the vehicle’s major systems, such as the engine or transmission. Sensors on each truck collect streaming IoT data in real time to provide context. This data includes the location of the mechanical event as well as the conditions present during the failure, such as altitude, ambient air temperature, RPM level, and torque load. The goal is to provide faster, more comprehensive diagnostics and repairs than traditional “set mileage schedule” service appointments, while providing fleet operators with information about driving habits, fuel consumption, truck performance and other factors.
Over the past few years, Volvo Trucks has been focused on improving and expanding its analytical capabilities. It partnered with analytics vendor SAS to deploy a more advanced platform that leverages event stream processing and artificial intelligence. “The SAS Analytics Platform provides Volvo Trucks with a way to define complex rules to improve how its remote diagnostic cases are created and operated,” said Deedy.
Data destination
Every Volvo truck equipped with remote diagnostics in North America has hundreds of sensors, about 75 of which are closely monitored because analysis determined they provide useful information about critical malfunctions. The truck uses the wireless cellular network to send data to the telematics communications gateway.
The proprietary gateway “is designed to access the core data network within the truck to reliably collect and store data without disrupting truck operations,” Deedy said. Critical data is transmitted directly to Volvo Trucks’ dedicated servers, which transform the information for further processing.
Any data indicating a mechanical failure is sent in near real-time to an inbound message queue, where it is extracted and processed by the SAS event stream processing engine over the 10Gb network.
The SAS system applies analytics-based rules and can trigger the creation of support cases through an outbound message queue. Volvo Trucks’ ASIST system is a web-based service management platform that pulls messages from outbound message queues to handle customer notifications.
“The rules engine analyzes complex business rules and applies them to streaming fault data in motion,” Deedy said. Volvo Trucks Uptime Center employees will then notify the customer of the fault via email, text message or phone call. “All data collection and processing is done within Volvo Trucks’ firewall to ensure data security.”
The system processes 1.5 million fault alerts per day, each containing 30 to hundreds of data items, such as powertrain sensor readings. SAS’s event stream processing triggers approximately 4,000 rules per day, including approximately 3,000 updated cases and 1,000 new cases.
Another component of the system, SAS Asset Performance Analytics, is the analytics engine Volvo Trucks uses to examine historical failure data. Administrators can use analytics capabilities to refine parameters in the rules engine, perform data selection, and report historical failure data.
The analytics engine stores 8 TB of data and saves historical failure data from 2015 to the present. These failures, which now total approximately 1 billion, were generated during this period by the growing fleet of connected trucks (currently 350,000 vehicles). 1.5 million faults are added to the analysis engine every day.
Due to operational complexity, the SAS platform took approximately 11 months to deploy and is built on high-performance server clusters using in-memory technology and low-latency/high-throughput solid-state drives.
Artificial intelligence capabilities enable Volvo Trucks to discover hidden insights in collected data and combine it with its engineering team’s truck knowledge. This enables companies to better understand the exact meaning of the data and integrate it into remote diagnostic services. SAS analytics deployment significantly reduces false alarms and improves diagnostic accuracy for Volvo Trucks.