Offline driving pattern detection and identification under usage disturbances

Optimizing the configuration of a wheel loader to customer needs can lead to a significant increase in efficiency with respect to fuel economy, cost, component dimensioning etc. Experience show that even modest customer adaptation can save around 20% of fuel cost. A key motivator for this work is that wheel loader manufacturers in general does not have full information about customer usage of the machine and the main objective here is to develop an algorithm that automatically, using only production sensors, extracts information about the usage of a machine at a specific customer site. Two main challenges are that sensors are not located with respect to this task and the significant usage disturbances that typically occur during operation. The proposed solution is a robust method, based on a mix of techniques using basic signal processing, state automaton techniques, and parameter estimation algorithms. A key property of the method is the method of combining, individually very simple, basic techniques in a scheme where robustness are introduced. The approach is evaluated on measured data of a wheel loader loading gravel and shot rock.