Eran Friedman - Boosting simulation performance with Python

Boost simulation performance with Python by leveraging a custom-built simulator powered by the CMPI library, enabling real-time testing, distributed simulation, and parallel processing for complex systems.

Key takeaways

The time command shows the time it took for the robot to run, but doesn’t display any messages in the screen, which is the problem.
SymPy has support for event-driven processes but runs all processes in a single thread, so it’s not suitable for the situation.
To solve the problem, all components should be tied to the clock.
The simulator maintains its own clock and immediately moves from one event to the next, allowing faster simulation.
The simulator was used to run the system in real-time and test the code before running it on the real robots.
The simulation helped to identify bugs that were hard to see in the real system, and allowed for easier testing.
The simulator was also used to evaluate the new warehouse layout before it was built.
The simulation improved the performance of the system and allowed for easier debugging.
The simulator supports distributed simulation, which allows for testing of complex systems.
The system uses a message queue to communicate between different parts of the system.
The simulator runs the system in a multi-process simulation, which allows for more realistic simulations.
The simulator uses a shared time tick to synchronize the different processes, and prevents one process from running too fast.
The simulator was implemented using the CMPI library, which provides a framework for discrete event simulation.
The system uses a Delia service to manage the time and prevent one process from running too fast.
The simulator is able to run the system in a distributed way, which allows for more realistic simulations of complex systems.
The simulator has several benefits, including improved performance and easier debugging.
The simulator can also be used to evaluate the system under different scenarios, such as different layouts or traffic patterns.
The simulator can run the system in a multi-threaded way, which allows for more realistic simulations of complex systems.
The simulator supports real-time simulation, which allows for real-time testing of the system.
The simulator can be used to evaluate the system in real-time, which allows for more accurate testing.
The simulator has been used to evaluate the system in real-world scenarios, such as in a warehouse with real robots.

Eran Friedman - Boosting simulation performance with Python

More talks