Questions Regarding Aggressor Workload and ROS1 Dependency

[Yinchen Ni]

Dear Arm Industrial Challenge Coordinators,

I am a researcher from Shanghai Jiao Tong University, and I’m very interested in contributing to the AR HUD case study presented in the ECRTS 2022 Industrial Challenge.

After carefully reviewing the official PDF and available resources, I have encountered a few issues and would greatly appreciate your guidance:

1. The paper mentions that synthetic implementations for aggressor workloads will be provided. However, I could not locate them in any of the referenced GitHub repositories. Could you please clarify where these synthetic workloads (or their models) can be obtained?

2. The OV2SLAM implementation relies on ROS1, which will reach EOL in May 2025. Moreover, it is increasingly difficult to set up ROS1 on modern systems such as Ubuntu 20.04, 22.04, or 24.04, due to compatibility and packaging issues.

3. Due to the above ROS1 dependency, it becomes challenging to run the reference model and obtain meaningful execution time measurements for the tasks, which is critical for the performance analysis and modeling part of the challenge.

Would you consider providing:

• Updated profiling data or representative execution traces for the major tasks?

• A pointer to any discussion or ongoing effort related to ROS2 migration, or alternative ways to measure task behavior?

I appreciate your time and thank you for organizing such a valuable challenge. Looking forward to your reply.

Best regards,
Yinchen