Code and result tables for "Adaptive UAV traffic monitoring under partial and delayed sensing: when does the expected value of surrogate-risk information change dispatch?"

Abstract

Scripts, configurations and result tables for the article "Adaptive UAV traffic monitoring under partial and delayed sensing: when does the expected value of surrogate-risk information change dispatch?" by A. Mahmoodi, M. Davoodi, E. A. Torkamani and S. M. Sajadi, submitted to Transportation Research Part C: Emerging Technologies. The archive holds a chronological replay of pNEUMA drone trajectories (Athens, October 2018) that compares information-valued (EVRI) and risk-greedy dispatch of one to four UAVs under partial and delayed ground sensing, in two areas. It contains the pipeline scripts, the ground-sensing scenario configurations and flight graphs, the fitted measurement updates, and the result tables reported in the article. The README inside the archive describes the layout, the order of the pipeline and the software used. Not included: the pNEUMA trajectory files, which are available at https://doi.org/10.5281/zenodo.10491409; the derived trajectories; and the per-run replay logs, which are available from the corresponding author. Road geometry extracts are from OpenStreetMap (© OpenStreetMap contributors) and remain under the ODbL. Use of AI tools: the scripts in this archive were prepared with a generative AI tool (Claude, Anthropic) under the authors' direction; the authors reviewed them and take responsibility for them. Their outputs were checked with deletion audits, unit tests and a reproduction check, as described in the article. The file outputs/phase4_visual_audit/audit_results.json contains first-pass labels of video clips that were produced with the same tool and then confirmed by two of the authors.