Why CARLA Is Not the Digital Twin in My Project

When the term digital twin is mentioned in autonomous vehicle research, many people assume it refers to a simulation environment. In many cases, this leads to the assumption that the CARLA simulator itself acts as the digital twin.

However, in my project this interpretation is not entirely accurate.

CARLA plays an important role in the development process, but it is not the digital twin.

Understanding this distinction was one of the most important conceptual steps in the project.

What CARLA Does

CARLA is an open-source autonomous driving simulator designed to support research in perception, planning, and control for autonomous vehicles [1].

The platform provides realistic urban environments with configurable sensors such as cameras, LiDAR, and radar. These features allow researchers to evaluate perception algorithms and autonomous behaviour under controlled conditions.

In this project, CARLA was used during the early development phase to test perception algorithms and evaluate the collision detection pipeline.

This allowed the system to be developed safely before integrating real sensors.

What the Digital Twin Actually Is

The digital twin in this project is not the simulator itself.

Instead, it is the software world model that represents the environment surrounding the vehicle.

This internal model stores information such as:

• detected obstacles

• object positions

• distance measurements

• velocity estimates

• collision risk values

Every time new sensor data arrives, the world model updates.

This makes the digital twin a continuously evolving representation of the real environment.

Linking Simulation and Reality

One of the key architectural ideas in the project is that the same perception and reasoning system can operate using both simulated and real sensor inputs.

During development, CARLA provides simulated sensor data.

During hardware experiments, the real sensors provide the data instead.

CARLA Sensors
        ↓
Perception Pipeline
        ↓
Digital Twin (World Model)
        ↓
TTC Risk Evaluation
Real Sensors
(Camera + LiDAR + Radar)
        ↓
Perception Pipeline
        ↓
Digital Twin (World Model)
        ↓
TTC Risk Evaluation

Because the digital twin remains independent of the data source, the system can transition from simulation to real hardware without changing its core architecture.

Why This Distinction Matters

Simulation and digital twins serve different purposes.

Simulators generate environments that allow algorithms to be tested safely.

Digital twins represent and interpret the state of a real system using live data streams [2].

In this project, CARLA is used as a development and testing tool, while the digital twin acts as the central reasoning layer that evaluates collision risk.

References

[1] A. Dosovitskiy et al., “CARLA: An Open Urban Driving Simulator,” Proceedings of the Conference on Robot Learning, pp. 1–16, 2017.

[2] S. Tao, Q. Qi, A. Liu and A. Kusiak, “Digital Twins and Cyber–Physical Systems toward Smart Manufacturing and Industry 4.0,” Engineering, vol. 5, no. 4, pp. 653–661, 2019.

[3] M. Ali, S. Hossain and M. Rahman, “Digital Twin in Intelligent Transportation Systems: A Review,” Applied Sciences, vol. 13, no. 10, 2023.