Generates a dataset for training on a SLURM cluster. Each route file is parallelized on its own machine. Monitors the data collection and continues crashed processes. Best run inside a tmux terminal.