Poster Presentation: Multi-Layer Digital Twin Development for Electric Trucks

Digital Twins (DTs) are virtual models that are capable
of bi-directional data exchange with their physical counterparts, ensuring synchronization, high accuracy, real-time performance, and
scalability. This work presents an advanced DT framework for electric trucks, featuring a universal six-layer DT architecture consisting of physical space, a communication channel, and a digital twin
space. The DTs are designed to be adaptive, self-calibrating, and
real-time prediction capable, enabling process optimization, monitoring, and maintenance. The objective is to enhance energy efficiency in medium freight haulage trucks by 10% through improved
vehicle operation, logistics planning, and charge scheduling. The
DTs will be applied to demonstration trucks for six months with at
least 200 km of daily driving. There are two use cases with differences in driving profiles and weight categories: • UC1: Distribution
Logistic (20t) • UC2: Refuse Collection (16t)
The following DTs are developed to satisfy the objective in compliance with the multi-layer framework: • Eco-driving DT optimizes
energy efficiency and extends driving range by estimating the
optimal speed profile to minimize consumption per distance
traveled. Simulation results for a 174 km trip show an SoC
usage of 39• Thermal System DT monitors temperature states at
specific points of interest to enhance control strategy and reduce
energy consumption. Simulation results compare two models, a
reduced-order empirical model and a data-driven model (LSTM).
The reduced-order model showed better accuracy in temperature
peak detection and was selected for the DT. • Energy Consumption DT calculates the predicted energy consumption for a route.
Simulation results show an error of 3% for the prediction of energy
consumption. Future work will integrate the Thermal System DT
to further improve the outcome. • Battery Ageing DT predicts
capacity degradation for an input operating condition. Simulation
results show the ability of the DT to capture the effect of the various
stress factors but validation could not be performed due to lack of
ageing data. • Charge Time Estimation DT generates the charging
power profile and predicts the charging time for given conditions.
Prediction error for data collected from an in-flied vehicle was
found to be within 2%. • Multi-level Control System Optimization
DT optimizes the smart vehicle control systems at component,
sub-system, vehicle and cloud levels. The algorithm has been
tested for feasibility of execution for the vehicle embedded platform.