Page 29 - Commercial Vehicle Engineer - June 2021

PREDICTIVE MAINTENANCE
“Predictive maintenance will reduce drastically the amount of unscheduled stops and roadside breakdowns as we can intervene earlier”
“The software is smart as it starts a chain
reaction, so when predicting required stock levels,
it automatically initiates a replenishment request with the re-ordered parts then delivered without
any human involvement. In some cases, this is
the level of automation that is wanted, in others
our customers also apply the order authorisation processes that they use for manually generated orders. The level of saving that can be achieved varies on a case-by-case basis but some of our customers have reported a reduction by 30%.”
In addition, the software can highlight where workshops are – or aren’t – getting value for money on their parts purchasing. “Many workshops just order the same part from the same supplier year-in year-out,” notes Tandy. “Analysis of historical data can reveal poor products. Using tablets, engineers can immediately see the vehicle’s maintenance history and spot recurring defects. Freeway can also pool data from multiple locations, gaining, for example,
an insight into variations in the prices paid for a particular part from different suppliers. Again, the software can learn and then alert managers if they are paying too much when signing off a purchase order.”
New trucks provide a wealth of data that technicians can use to plan maintenance
Technicians can see what parts are needed at the tap of a tablet
Artificial intelligence
As Tandy mentions, the software provided by Freeway, in common with other providers, is smart. “Using artificial intelligence – or ‘machine leaning’ – to generate analytic frameworks the software now presents large data sets in formats that are readily understandable and provides
the information needed for management action. The maintenance record data utilised to generate these approaches has been stored in the Freeway fleet management system from the inception of electronic data entry. [Some customers have] used the system for decades and there are many hundreds of years of data that provides the foundations for the approach that has been taken.
“This system also helps to identify recurring trends and anomalies such as fluctuating seasonal demand and likely extra demand for particular vehicles scheduled for servicing. The software continually learns from historic data to create
a more accurate predictive model by tracking performance of all assets at the level of individual components so that entire-life costs of individual vehicles as well as groups of vehicles by make and model or type of asset or user-defined dimension can be understood.”
But it isn’t just about automated learning, Tandy adds. “The software has also benefited from many years of ‘human learning’ over the last 25 years with our customers’ requests and suggestions equally adding to the utility of the data frameworks.”
COMMERCIAL VEHICLE ENGINEER > JUNE 2021 29