Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in Coakley et alK. L. Coakley, T. Snelleman, H. Hoos, and O. E. Gundersen, "The embrace of open science: An analysis of a decade of AI research and 56 800 conference papers," Under Review, 2026..
Convergent Plans for Large-Scale Evacuations
Authors: Caroline Even, Victor Pillac, Pascal Van Hentenryck
AAAI 2015 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experimental results on a real case study show that the two-stage approach produces better primal bounds than the MIP model and is two orders of magnitude faster; It also produces dual bounds stronger than the linear relaxation of the MIP model. Finally, simulations of the evacuation demonstrate that convergent evacuation plans outperform existing approaches for realistic driver behaviors. |
| Researcher Affiliation | Academia | 1National ICT Australia (NICTA), Melbourne, Australia 2Australian National University (ANU), Canberra, Australia |
| Pseudocode | Yes | Algorithm 1 The two-stage Algorithm TDFS for the CEPP; Algorithm 2 The two-stage Algorithm TDFS-CT for the CEPP with Minimization of the Clearance Time |
| Open Source Code | No | The paper does not contain any statement about releasing source code or provide a link to a code repository. |
| Open Datasets | No | The paper mentions using data from a 'real case study' of the Hawkesbury-Nepean (HN) ๏ฌoodplain and describes its characteristics (nodes, edges, time horizon, scaled instances) but does not provide a link, DOI, or a citation to a publicly accessible version of this specific dataset. |
| Dataset Splits | No | The paper does not specify explicit training, validation, or test dataset splits. It evaluates on different problem instances (HN80-Ix) but does not detail data partitioning for model training or validation. |
| Hardware Specification | Yes | The algorithms were implemented using JAVA 7 and GUROBI 5.6 and the results were obtained on 64bits machines with 2.8GHz AMD 6Core Opteron 4184 and 32Gb of RAM. |
| Software Dependencies | Yes | The algorithms were implemented using JAVA 7 and GUROBI 5.6 |
| Experiment Setup | Yes | The time horizon H spans 10 hours discretized into 5 minutes time-steps. For each iteration of TDFS, we allocate 30s for each call to TDP. ...the TDP-CT model is stopped after 30s when the duality gap is lower than 1%. |