A Path through the Technician and Driver Shortages: Predictive Maintenance

On-highway fleets have two perennial challenges in driver and technician retention.

The good news is that carriers already have a profitable path forward in the data they capture from their vehicles and maintenance. The key to starting down that road is the value and efficiency brought to fleet management through predictive maintenance.

In a recent report, the American Trucking Association (ATA) figured the industry’s shortage of drivers at 80,000. CDL schools and apprentice programs are not attracting or turning out enough new drivers. In turn, fleets have eyed nontraditional sources of recruits to shore up their shortages, including foreign workers and teenagers.

New recruits can slow the revolving door, but don’t address one of the primary causes of driver frustration and turnover. If issues aren’t repaired efficiently, and drivers miss out on deliveries, then both the repair shop and driver retention issues jeopardize deliveries and business continuity. Lapses in maintenance compound into unexpected downtime and roadside breakdowns, and undercut driver pay. The annual turnover for large truckload carriers is hovering at 90%. For smaller fleets, it is about 70%.

Fleets are seeing similar challenges with the recruitment of technicians. Just to keep pace with the growing maintenance demands in the industry, the U.S. Bureau of Labor Statistics estimates that the industry would need to add 75,000 new diesel technicians by 2024. However, technical schools are only graduating 3,500 diesel technicians each year.

Once technicians break into the industry, many are not staying. A survey from Automotive Survey Excellence (ASE), the accrediting body for diesel technicians for high schools and post-secondary institutions, reported that 42% of new diesel technicians left the field within two years.

While the shortage of technicians is cause for concern, it also underscores a significant challenge that fleets are facing: the gap in technician skills. With the emergence of new alternative vehicles and the growing dependence on onboard computers, technical expertise today is undergoing rapid change. In addition to mastering these emerging technologies, new technicians are entering repair shops that are expecting experienced diesel technicians to soon retire.

With challenges to driver and technician staffing, carriers are running up maintenance costs and scrambling to meet their delivery obligations.

The ATA estimates that the cost to hire a new driver can be at least $8,000, which doesn’t yet account for the costs of onboarding as the driver adjusts to new in-cab specs. Open technician roles can be costly as well –– up to $1,200 in productivity for a repair shop.

Taking care of these problems is especially important now. While viewed traditionally as a cost center, efficient fleet maintenance fosters reliability and opportunities for business development. And as consumer demand shifts from consumer goods to alternatives like entertainment and the industry deals with a delayed procurement schedule for new vehicles, an expensive aftermarket, and overall supply chain disruptions, efficient maintenance can be a cushion for the bottom line.

While fleets might be short-handed and feeling the brunt of these shortages, they have another resource on staff to make maintenance more efficient: vehicle performance data. And fortunately for fleets, the data is captured and collected in existing fleet management technology like computerized maintenance management systems (CMMS) and enterprise asset management (EAM) software.

These software systems usually contain large volumes of data. They work great as repositories for data. But as any fleet manager knows, large volumes of data quickly become unwieldy. Too much information becomes overwhelming and, at a certain point, unhelpful. As a result, fleet managers fall back on time-honored customs for reactive and preventive maintenance, whether or not those practices align with the best interests of a fleet or its customers.

The issue isn’t in the collection of data in these systems, but in making organized sense of it. Fault codes, sensors data, fluid analysis readings, driver performance data, work orders, and dispatch information each provide a glimpse into vehicle conditions and maintenance.

With all this data and running the repair shop, fleet managers lack the time or bandwidth to parse all of that information. They certainly could, and have the subject matter expertise to do so, but the ability to prioritize is the great advantage of technology like predictive maintenance to automatically sort through the data. It gives relevant information to overworked and overstretched maintenance teams so that repair shops can focus on their turning wrenches.

Predictive analytics for vehicle maintenance can be thought of as a smarter filter, where using CMMS or EAM systems are a dragnet. Predictive maintenance sifts through existing data to bring out impactful answers to the top for maintenance teams.

Often, predictive analytics process the data through data science models that understand the differences in data for a given system or subsystem in a vehicle. For example, the model might be designed to detect, diagnose, and alert on issues with engine overheating.

Models train on years of vehicle data associated with engine overheating, learning the difference between ideal and abnormal performance of the failure and pinpointing its causes. Since vehicles run in different terrains and conditions at different duty cycles, contextual sources of information are important for tuning the models to generate precise predictions on future engine and vehicle performance.

Just as important as the context of data is to the training of the data science model, the context of the data science model to the fleet’s profitability is significant. That is to say, not all impending vehicle issues are created equal. It’s important for models to distinguish the severity of potential issues for any vehicle. And for a fleet-wide view, it’s important for predictive analytics to create a quick checklist of issues to be addressed, and how that order of maintenance operations fits on to overall fleet objectives.

The outcome of an individual data science model puts one technician in a better position to fix one issue. Across a fleet, predictive maintenance puts carriers in a position to be proactive on vehicle diagnostics, repairs, utilization, and load assignments. Fleet managers have the direct support to make more profitable decisions within the course of normal maintenance operations.

As the ranks of technicians thin and repair shops continue to see vehicles roll in, the effect of extended downtime on driver retention will continue to be felt. For many carriers, this relationship is familiar. And while private fleets generally have stronger records with driver retention, the promise of more revenue miles can push even the most loyal drivers away to competitors.

With more notice of impending failures and the ability to plan more repairs, predictive maintenance software puts fleets in a position to make better decisions that result in greater uptime –– from load assignment to diagnostics and repair.

Starting with load assignments, predictive maintenance enables fleets to deploy route-ready assets. Vehicles flagged with critical issues can be pulled from service and tended to in the shop, sparing the shop an expensive repair and the fleet (and driver) valuable time on the road.

The advanced visibility also enables fleets to reduce routine sources of frustration for drivers. By ensuring greater reliability, predictive maintenance opens up more regularity in vehicle utilization and limits slip-seating. And then there is the reduction of unplanned downtime that cuts into their paychecks and eats away at their patience.

While predictive maintenance provides real-time visibility into vehicle conditions, it’s not always possible to carry out repairs in a proactive manner. Still, the availability of insights in real-time allows fleets to make over-the-air adjustments on critical failures. Fleet managers have an open window to call an asset into an in-house shop or the dealership and make the repair that avoids a roadside breakdown and a tow. Predictive maintenance provides the assurance that the fleet can catch catastrophic failures before they happen.

For repairs, predictive maintenance analytics provide insight into conditions previously unknown to the fleet. It’s like having another seasoned technician on staff, but one that is focused squarely on the most expensive mechanical problems to a fleet –– issues like diesel particulate filter (DPF) failures and cylinder head failures for compressed natural gas (CNG) vehicles. These routine issues depend on the skills of an experienced technician to spot. With predictive maintenance, fleet managers and technicians receive automated and proactive notice.

Predictive maintenance benefits fleet maintenance, operations, and drivers. Still, maintenance has historically been treated as an expense the business sinks into operations. For private fleets, this emphasis on maintenance as an expense rings especially true. Time in the shop making repairs equals time lost out on the road.

As any fleet manager knows, less maintenance isn’t necessarily a fleet’s best bet. Vehicle utilization without due inspections or repairs is a wager against time and reliability. Planned, well-timed repairs that fit business development and bottom-line goals are easier to control with predictive maintenance. Even as issues with driver and technician retention make their mark on new revenue opportunities and the maintenance budget, predictive maintenance offers fleets the opportunity to realize the value of their data and make better decisions at scale.

Ready to put your data to work?