efleetplan 1.3__tar.gz

efleetplan-1.3/.gitignore

@@ -0,0 +1,7 @@
+__pycache__/
+*.pyc
+*.egg-info/
+dist/
+build/
+.venv/
+venv/

efleetplan-1.3/LICENSES/License.txt

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Carolina Gil Ribeiro
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

efleetplan-1.3/PKG-INFO

@@ -0,0 +1,227 @@
+Metadata-Version: 2.5
+Name: efleetplan
+Version: 1.3
+Summary: Co-optimisation tool of charging infrastructure investment and electric fleet operations. EFleetPlan is a Python-based tool for optimizing charging infrastructure and operations for electric light commercial vehicles (LCVs).
+Project-URL: Repository, https://github.com/mcarolinagilr/efleetplan
+Project-URL: Issues, https://github.com/mcarolinagilr/efleetplan/issues
+Author: KTH Royal Institute of Technology
+Author-email: Carolina Gil Ribeiro <mcgr2@kth.se>
+License-Expression: MIT
+License-File: LICENSES/License.txt
+Keywords: EV,LCV,Pyomo,charging,fleet,infrastructure,optimization
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Science/Research
+Classifier: Operating System :: OS Independent
+Classifier: Programming Language :: Python :: 3
+Requires-Python: >=3.10
+Requires-Dist: matplotlib<4.0,>=3.10
+Requires-Dist: numpy<3.0,>=2.3
+Requires-Dist: pandas<3.0,>=2.3
+Requires-Dist: pydantic<3.0,>=2.0
+Requires-Dist: pyomo<7.0,>=6.9
+Requires-Dist: pyyaml<7.0,>=6.0
+Description-Content-Type: text/markdown
+
+# eFleetPlan
+Co-optimisation tool of charging infrastructure investment and electric fleet operations
+
+**EFleetPlan** is a Python-based open source tool for optimizing charging infrastructure and operations for electric light commercial vehicles (LCVs).
+
+## Authors
+Carolina Gil Ribeiro and Jagruti Thakur
+
+## License
+This software is licensed under the MIT License.  
+See the [LICENSE](./LICENSE) file for details.
+
+
+## Citation
+If you use eFleetPlan, please cite:
+**Gil Ribeiro, C and Thakur, J, eFleetPlan: Co-optimisation tool of Charging Infrastructure Investment and Fleet Operations, 2025. DOI:xxxxxxx
+
+## Project structure
+
+```
+eFleetPlan/
+├── config/                          # All configuration files
+│   ├── predefined/                  # Predefined parameter sets
+│   │   ├── companies.yaml
+│   │   ├── infrastructure_configuration.yaml
+│   │   ├── schedules.yaml
+│   │   └── vehicles.yaml
+│   ├── env.yaml                     # Environment settings 
+│   ├── run_FleetSchedule_Config.yaml# Schedule generation run configuration
+│   └── run_Optimisation_Config.yaml # Optimisation run configuration
+├── data/
+│   ├── Input/                       # Energy consumption and electricity price files
+│   └── Output/                      # Results files
+├── docs/                            # Documentation
+├── notebooks/
+│   ├── 1_Fleet_Operation_simulation.ipynb
+│   └── 2_Co-optimisation.ipynb
+├── src/
+│   └── efleetplan/
+│       ├── _1_schedule/             # Fleet operation simulation package
+│       │   ├── generate_graphs.py
+│       │   └── schedule_generation.py
+│       └── _2_optimisation/         # Co-optimisation package
+│           ├── co_optimisation.py
+│           └── optimisation_graphs.py
+├── pyproject.toml
+├── MANIFEST.in
+└── mkdocs.yml
+```
+
+## Installation instructions
+
+### Prerequisites
+
+- Python 3.9 or later
+- Solver for optimisation: default is [Gurobi Optimizer](https://www.gurobi.com/) with a valid license but can be changed for other solver in the co_optimisation.py file 
+
+There are two ways to install eFleetPlan, depending on what you want to do.
+
+### Option A — From GitHub (clone the repo)
+
+Use this to run the example notebooks, reproduce the paper's illustrative examples, or edit the source code.
+
+**Step 1 — Clone the repository**
+
+```bash
+git clone https://github.com/mcarolinagilr/efleetplan.git
+cd efleetplan
+```
+
+**Step 2 — Create a virtual environment and install**
+
+```bash
+python -m venv venv
+source venv/bin/activate      # Windows: venv\Scripts\activate
+pip install .
+```
+
+**Step 3 — Verify the installation**
+
+```bash
+python -m efleetplan
+```
+
+This confirms that all dependencies (including Gurobi) are correctly installed.
+
+### Option B — From PyPI (use as a library)
+
+Use this to call eFleetPlan's functions from your own code, without cloning the repository.
+
+**Step 1 — Create a virtual environment and install**
+
+```bash
+python -m venv venv
+source venv/bin/activate      # Windows: venv\Scripts\activate
+pip install efleetplan
+```
+
+**Step 2 — Create your config folder**
+
+```bash
+efleetplan-start
+```
+
+Prompts for a destination folder and copies editable YAML templates there (`env.yaml`, run configs, the predefined vehicle/schedule/company/infrastructure library, and the illustrative example configs). Edit those, then supply your own energy consumption factor and electricity price CSVs.
+
+**Step 3 — Use it**
+
+```python
+from efleetplan import load_scheduler_config, generate_fleet_schedules
+
+env, run, predefined = load_scheduler_config(
+    env_yaml="config/env.yaml",
+    run_yaml="config/run_FleetSchedule_Config.yaml",
+)
+schedule = generate_fleet_schedules(env, run, predefined)
+```
+
+## eFleetPlan Configurations
+
+**All parameters are configured through YAML files in the `config/` directory. There is no need to edit the notebooks or Python source code.**
+
+### Environment settings — `env.yaml`
+Defines global settings such as file paths, simulation period, time resolution, and the random seed.
+
+### Schedule generation — `run_FleetSchedule_Config.yaml`
+
+Controls how fleet operational schedules are generated: the schedule name, number of vehicles, schedule and vehicle mix, and company type. 
+User can use predefined parameters or define custom values.
+
+Predefined values are loaded from `config/_1_predefined/`.
+
+### Optimisation — `run_Optimisation_Config.yaml`
+
+Defines settings for the Co-optimisation model: which schedule to optimise, fleet size, solver gap tolerance, and infrastructure configuration (Cost and power parameters). User can use predefined parameters or define custom values. 
+
+Predefined values are loaded from `config/_1_predefined/infrastructure_configuration.yaml`.
+
+
+### Predefined parameter sets — `config/_1_predefined/`
+
+Contains reusable definitions for vehicles, schedules, companies, and infrastructure. These can be extended with new entries as needed.
+
+### Input data — `data/Input/`
+
+Place the following files in this folder before running the tool:
+
+- **Energy consumption factor file** — energy consumption factor, that is linked to the environment temperature.
+- **Electricity price file** — time-series electricity prices for the simulation period.
+
+
+
+## How to use the eFLeetPlan tool
+
+EFleetPlan is run through two Jupyter notebooks. Both notebooks load their configuration from the YAML files described above, so all parameter changes should be made there before launching.
+
+### Fleet Operation simulation
+
+**Notebook:** `notebooks/1_Fleet_Operation_simulation.ipynb`
+
+Generates operational schedules (travel and charging patterns) for an electric LCV fleet based on the configured parameters. The notebook validates inputs, runs the generation, and produces visualisations of the resulting schedules.
+
+**Workflow:**
+1. Edit `config/env.yaml` and `config/run_FleetSchedule_Config.yaml` with your desired settings.
+2. Open the notebook and run all cells.
+3. Review the validation checks and output graphs.
+4. Outputs are saved to `data/Output/<schedule_name>/`.
+
+### Package 2 — Charging Infrastructure Co-optimisation
+
+**Notebook:** `notebooks/2_Co-optimisation.ipynb`
+
+Solves the cost-minimisation model that jointly optimises charging infrastructure investment and fleet charging operations. Requires Gurobi and a schedule generated by Package 1.
+
+**Workflow:**
+1. Edit `config/run_Optimisation_Config.yaml` with the target schedule and solver settings.
+2. Open the notebook and run all cells.
+3. Review the optimisation results and summary visualisations.
+4. Results are saved to `data/Output/<schedule_name>/Results/`.
+
+### Running the notebooks
+
+```bash
+# From the project root
+jupyter lab
+# or
+jupyter notebook
+```
+Open the notebook for the Package you want to run and execute cells from top to bottom.
+
+## Output
+
+Package 1 produces CSV files with the generated fleet schedules and accompanying graphs in `data/Output/<schedule_name>/`.
+
+Package 2 produces several result files in `data/Output/<schedule_name>/Results/`, including the main decision variables, per-vehicle results, summary tables, and hourly aggregated statistics (averages and maxima).
+
+
+
+
+
+
+
+

efleetplan-1.3/README.md

@@ -0,0 +1,203 @@
+# eFleetPlan
+Co-optimisation tool of charging infrastructure investment and electric fleet operations
+
+**EFleetPlan** is a Python-based open source tool for optimizing charging infrastructure and operations for electric light commercial vehicles (LCVs).
+
+## Authors
+Carolina Gil Ribeiro and Jagruti Thakur
+
+## License
+This software is licensed under the MIT License.  
+See the [LICENSE](./LICENSE) file for details.
+
+
+## Citation
+If you use eFleetPlan, please cite:
+**Gil Ribeiro, C and Thakur, J, eFleetPlan: Co-optimisation tool of Charging Infrastructure Investment and Fleet Operations, 2025. DOI:xxxxxxx
+
+## Project structure
+
+```
+eFleetPlan/
+├── config/                          # All configuration files
+│   ├── predefined/                  # Predefined parameter sets
+│   │   ├── companies.yaml
+│   │   ├── infrastructure_configuration.yaml
+│   │   ├── schedules.yaml
+│   │   └── vehicles.yaml
+│   ├── env.yaml                     # Environment settings 
+│   ├── run_FleetSchedule_Config.yaml# Schedule generation run configuration
+│   └── run_Optimisation_Config.yaml # Optimisation run configuration
+├── data/
+│   ├── Input/                       # Energy consumption and electricity price files
+│   └── Output/                      # Results files
+├── docs/                            # Documentation
+├── notebooks/
+│   ├── 1_Fleet_Operation_simulation.ipynb
+│   └── 2_Co-optimisation.ipynb
+├── src/
+│   └── efleetplan/
+│       ├── _1_schedule/             # Fleet operation simulation package
+│       │   ├── generate_graphs.py
+│       │   └── schedule_generation.py
+│       └── _2_optimisation/         # Co-optimisation package
+│           ├── co_optimisation.py
+│           └── optimisation_graphs.py
+├── pyproject.toml
+├── MANIFEST.in
+└── mkdocs.yml
+```
+
+## Installation instructions
+
+### Prerequisites
+
+- Python 3.9 or later
+- Solver for optimisation: default is [Gurobi Optimizer](https://www.gurobi.com/) with a valid license but can be changed for other solver in the co_optimisation.py file 
+
+There are two ways to install eFleetPlan, depending on what you want to do.
+
+### Option A — From GitHub (clone the repo)
+
+Use this to run the example notebooks, reproduce the paper's illustrative examples, or edit the source code.
+
+**Step 1 — Clone the repository**
+
+```bash
+git clone https://github.com/mcarolinagilr/efleetplan.git
+cd efleetplan
+```
+
+**Step 2 — Create a virtual environment and install**
+
+```bash
+python -m venv venv
+source venv/bin/activate      # Windows: venv\Scripts\activate
+pip install .
+```
+
+**Step 3 — Verify the installation**
+
+```bash
+python -m efleetplan
+```
+
+This confirms that all dependencies (including Gurobi) are correctly installed.
+
+### Option B — From PyPI (use as a library)
+
+Use this to call eFleetPlan's functions from your own code, without cloning the repository.
+
+**Step 1 — Create a virtual environment and install**
+
+```bash
+python -m venv venv
+source venv/bin/activate      # Windows: venv\Scripts\activate
+pip install efleetplan
+```
+
+**Step 2 — Create your config folder**
+
+```bash
+efleetplan-start
+```
+
+Prompts for a destination folder and copies editable YAML templates there (`env.yaml`, run configs, the predefined vehicle/schedule/company/infrastructure library, and the illustrative example configs). Edit those, then supply your own energy consumption factor and electricity price CSVs.
+
+**Step 3 — Use it**
+
+```python
+from efleetplan import load_scheduler_config, generate_fleet_schedules
+
+env, run, predefined = load_scheduler_config(
+    env_yaml="config/env.yaml",
+    run_yaml="config/run_FleetSchedule_Config.yaml",
+)
+schedule = generate_fleet_schedules(env, run, predefined)
+```
+
+## eFleetPlan Configurations
+
+**All parameters are configured through YAML files in the `config/` directory. There is no need to edit the notebooks or Python source code.**
+
+### Environment settings — `env.yaml`
+Defines global settings such as file paths, simulation period, time resolution, and the random seed.
+
+### Schedule generation — `run_FleetSchedule_Config.yaml`
+
+Controls how fleet operational schedules are generated: the schedule name, number of vehicles, schedule and vehicle mix, and company type. 
+User can use predefined parameters or define custom values.
+
+Predefined values are loaded from `config/_1_predefined/`.
+
+### Optimisation — `run_Optimisation_Config.yaml`
+
+Defines settings for the Co-optimisation model: which schedule to optimise, fleet size, solver gap tolerance, and infrastructure configuration (Cost and power parameters). User can use predefined parameters or define custom values. 
+
+Predefined values are loaded from `config/_1_predefined/infrastructure_configuration.yaml`.
+
+
+### Predefined parameter sets — `config/_1_predefined/`
+
+Contains reusable definitions for vehicles, schedules, companies, and infrastructure. These can be extended with new entries as needed.
+
+### Input data — `data/Input/`
+
+Place the following files in this folder before running the tool:
+
+- **Energy consumption factor file** — energy consumption factor, that is linked to the environment temperature.
+- **Electricity price file** — time-series electricity prices for the simulation period.
+
+
+
+## How to use the eFLeetPlan tool
+
+EFleetPlan is run through two Jupyter notebooks. Both notebooks load their configuration from the YAML files described above, so all parameter changes should be made there before launching.
+
+### Fleet Operation simulation
+
+**Notebook:** `notebooks/1_Fleet_Operation_simulation.ipynb`
+
+Generates operational schedules (travel and charging patterns) for an electric LCV fleet based on the configured parameters. The notebook validates inputs, runs the generation, and produces visualisations of the resulting schedules.
+
+**Workflow:**
+1. Edit `config/env.yaml` and `config/run_FleetSchedule_Config.yaml` with your desired settings.
+2. Open the notebook and run all cells.
+3. Review the validation checks and output graphs.
+4. Outputs are saved to `data/Output/<schedule_name>/`.
+
+### Package 2 — Charging Infrastructure Co-optimisation
+
+**Notebook:** `notebooks/2_Co-optimisation.ipynb`
+
+Solves the cost-minimisation model that jointly optimises charging infrastructure investment and fleet charging operations. Requires Gurobi and a schedule generated by Package 1.
+
+**Workflow:**
+1. Edit `config/run_Optimisation_Config.yaml` with the target schedule and solver settings.
+2. Open the notebook and run all cells.
+3. Review the optimisation results and summary visualisations.
+4. Results are saved to `data/Output/<schedule_name>/Results/`.
+
+### Running the notebooks
+
+```bash
+# From the project root
+jupyter lab
+# or
+jupyter notebook
+```
+Open the notebook for the Package you want to run and execute cells from top to bottom.
+
+## Output
+
+Package 1 produces CSV files with the generated fleet schedules and accompanying graphs in `data/Output/<schedule_name>/`.
+
+Package 2 produces several result files in `data/Output/<schedule_name>/Results/`, including the main decision variables, per-vehicle results, summary tables, and hourly aggregated statistics (averages and maxima).
+
+
+
+
+
+
+
+

efleetplan-1.3/config/_1_predefined/companies.yaml

@@ -0,0 +1,137 @@
+# =============================================================================
+# Company type/ Use-Case predefined
+# =============================================================================
+# Each key is a company type or commodity type, and the values are the parameters related to the distance and stops of the trips.
+#The values related to company type are defined based on the findings of a survey 
+#on light goods vehicles in Sweden conducted by Transport Analysis in 2022. 
+
+
+# Shared defaults 
+_defaults: &defaults
+  min_distance: 30
+  max_distance: 200
+  min_distance_per_step: 1
+  max_distance_per_step: 50
+
+
+# ----------------------------------------------------------------------------
+# Parameters per type of transport
+#-----------------------------------------------------------------------------
+
+# Type of transport is defined as the primary use of the LCV and are described as follows: 
+  # Distribution: Distribution transport for goods or commodities transports with several stops for loading and unloading along the way; 
+  # Linehaul: Line haul goods or commodities transports, directly from one place to another;  
+  # craft_serv_with_goods: Crafts and services with goods transports include trips for craft or service vehicle that include goods or merchandise to be used or installed in the work; 
+  # craft_serv_without_goods: Crafts and services without goods correspond to craft or service trips without goods or merchandise.
+
+
+distribution:
+  <<: *defaults
+  avg_distance_wd: 124
+  dev_distance_wd: 15
+  avg_distance_we: 73
+  dev_distance_we: 15
+  max_distance: 300            # override default
+  avg_stops: 52.4
+  dev_stops: 11.7
+
+linehaul:
+  <<: *defaults
+  avg_distance_wd: 73
+  dev_distance_wd: 16
+  avg_distance_we: 43
+  dev_distance_we: 16
+  avg_stops: 3.6
+  dev_stops: 1.1
+
+craft_serv_with_goods:
+  <<: *defaults
+  avg_distance_wd: 75
+  dev_distance_wd: 7
+  avg_distance_we: 45
+  dev_distance_we: 7
+  avg_stops: 52.4
+  dev_stops: 11.7
+
+craft_serv_without_goods:
+  <<: *defaults
+  avg_distance_wd: 71
+  dev_distance_wd: 8
+  avg_distance_we: 42
+  dev_distance_we: 8
+  avg_stops: 3.7
+  dev_stops: 0.5
+
+
+
+
+# ----------------------------------------------------------------------------
+# Parameters per comodity type
+# -----------------------------------------------------------------------------
+
+# Commodity groups were formulated in Light Goods Vehicles 2022 survey to capture the type of goods 
+# commonly transported identified by the vehicles.
+# Food: Food and beverages, including refrigerated transport;
+# Mail: Mail and parcels, including postal and courier services;
+# Building: Building materials, including construction and renovation materials;
+# High distance: Transports with high average distance, such as long-haul transport or transports;
+# General cargo: General cargo, including transports of goods or commodities not included in the other
+
+food:
+  <<: *defaults
+  avg_distance_wd: 81
+  dev_distance_wd: 21
+  avg_distance_we: 48
+  dev_distance_we: 21
+  avg_stops: 11
+  dev_stops: 4.1
+
+mail:
+  <<: *defaults
+  avg_distance_wd: 122
+  dev_distance_wd: 23
+  avg_distance_we: 72
+  dev_distance_we: 23
+  avg_stops: 98.1
+  dev_stops: 21.8
+
+building:
+  <<: *defaults
+  avg_distance_wd: 75
+  dev_distance_wd: 8
+  avg_distance_we: 44
+  dev_distance_we: 8
+  avg_stops: 4.1
+  dev_stops: 0.3
+
+highdistance:
+  <<: *defaults
+  avg_distance_wd: 350
+  dev_distance_wd: 50
+  avg_distance_we: 300
+  dev_distance_we: 50
+  min_distance: 30
+  max_distance: 400
+  max_distance_per_step: 70
+  avg_stops: 52.4
+  dev_stops: 11.7
+
+generalcargo:
+  <<: *defaults
+  avg_distance_wd: 108.6
+  dev_distance_wd: 23
+  avg_distance_we: 64
+  dev_distance_we: 23
+  min_distance: 20
+  avg_stops: 11.7
+  dev_stops: 3.3
+
+nogoods:
+  <<: *defaults
+  avg_distance_wd: 67.4
+  dev_distance_wd: 6.8
+  avg_distance_we: 39.9
+  dev_distance_we: 6.8
+  min_distance: 20
+  avg_stops: 3.4
+  dev_stops: 0.4

efleetplan-1.3/config/_1_predefined/infrastructure_configuration.yaml

@@ -0,0 +1,53 @@
+# =============================================================================
+# INFRASTRUCTURE CHARGING CONFIGURATIONS
+# =============================================================================
+
+
+# -----  INFRASTRUCTURE POWER ---------------------------------------------------------------------------
+
+Charging_losses: 0.964  # Charging efficiency
+
+Charger_Power: 
+  f1: 7.4
+  f2: 50
+  f3: 150
+  f4: 350
+  route: 150
+
+
+Battery_Maximum_Limit: 0.8  # Maximum limit of the battery (80% of its capacity)
+Battery_Minimum_Limit: 0.2   # Minimum limit of the battery (20% of its capacity)
+
+  
+# -----  INFRASTRUCTURE COSTS PARAMETERS---------------------------------------------------------------------------
+
+
+lifetime: 20 #years
+Infrastructure_subscription: 50 #SEK/day
+Price_FixedrateDT: 0.0331 #SEK/kWh
+Demand_rate: 1352 #kW
+Price_Fixedrateroute: 8.90 #SEK/route
+Discount_rate: 0.05 #%
+
+
+investment cost: 
+  f1: 32000 #SEK
+  f2: 294000 #SEK
+  f3: 778000 #SEK
+  f4: 1451000 #SEK
+  route: 778000 #SEK
+
+installation cost:
+  f1: 29000 #SEK
+  f2: 472000 #SEK
+  f3: 644000 #SEK
+  f4: 1202000 #SEK
+  route: 1202000 #SEK
+
+maintenance cost:
+  f1: 5000 #SEK/year
+  f2: 40000 #SEK/year
+  f3: 120000 #SEK/year
+  f4: 240000 #SEK/year
+  route: 120000 #SEK/year
+

efleetplan-1.3/config/_1_predefined/schedules.yaml

@@ -0,0 +1,52 @@
+# =============================================================================
+# Schedule Type Predefined
+# =============================================================================
+# Each top-level key is a schedule type name.
+# The Python config loader will map these directly to ScheduleConfig attributes.
+
+typea:
+  # Continuous schedule — no break in the middle of the day, just a single stop in the distribution terminal at the beginning and a single return at the end of the day
+  dep_mean_wd: 8
+  dep_dev_wd: 1
+  ret_mean_wd: 18
+  ret_dev_wd: 1
+  dep_mean_we: 8
+  dep_dev_we: 1
+  ret_mean_we: 18
+  ret_dev_we: 1
+  min_dep: 6
+  max_dep: 11
+  min_return_hour: 18
+  max_return_hour: 22
+
+typeb:
+  # Schedule with two parts, with a break /middle stop in the day at the distribution terminal for charging / reloading
+  dep_mean_wd: 6
+  dep_dev_wd: 1
+  min_dep: 3
+  max_dep: 10
+
+  pause_beg_mean_wd: 12
+  pause_beg_dev_wd: 0.25
+  pause_end_mean: 13
+  pause_end_dev: 0.25
+  max_beg_time: 13
+  min_beg_time: 11
+  max_pause_end: 15
+  min_pause_end: 12
+  pause_time_mean: 0.5
+  pause_time_dev: 0.1
+
+  ret_mean_wd: 19
+  ret_dev_wd: 1
+  max_return_hour: 22
+  min_return_hour: 15
+
+  dep_mean_we: 9
+  dep_dev_we: 1
+  pause_beg_mean_we: 12
+  pause_beg_dev_we: 0.25
+  ret_mean_we: 19
+  ret_dev_we: 0.5
+
+  prob_emergency: 0.02