diff --git a/.gitignore b/.gitignore
index ced047a..42120ce 100644
--- a/.gitignore
+++ b/.gitignore
@@ -1 +1,4 @@
-Enrichment
+Enrichment/
+__pycache__/
+Logs/
+[Dd]ata/
diff --git a/descriptors/__init__.py b/descriptors/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/descriptors/dfc_aps.py b/descriptors/dfc_aps.py
new file mode 100644
index 0000000..c632f9a
--- /dev/null
+++ b/descriptors/dfc_aps.py
@@ -0,0 +1,30 @@
+import numpy as np
+import yaml
+
+import sys
+sys.path.append('..')
+import ld2dap
+
+def run(rasters, treshold=1e4, areas=None, sd=None, moi=None):
+    # Parse attribute type
+    treshold = float(treshold)
+    areas = None if areas is None else np.array(areas).astype(np.float).astype(np.int)
+    sd = None if sd is None else np.array(sd).astype(np.float)
+    moi = None if moi is None else np.array(moi).astype(np.float)
+
+    # APs Pipelines
+    loader = ld2dap.LoadTIFF(rasters)
+    dfc_filter = ld2dap.Treshold(treshold)
+    dfc_filter.input = loader
+    aps = ld2dap.AttributeProfiles(area=areas, sd=sd, moi=moi)
+    aps.input = dfc_filter
+    out_vectors = ld2dap.RawOutput()
+    out_vectors.input = aps
+
+    # Compute vectors
+    out_vectors.run()
+    
+    return out_vectors.data
+
+def version():
+    return 'v0.0'
\ No newline at end of file
diff --git a/descriptors/dfc_base.py b/descriptors/dfc_base.py
new file mode 100644
index 0000000..dede241
--- /dev/null
+++ b/descriptors/dfc_base.py
@@ -0,0 +1,34 @@
+#!/usr/bin/python
+# -*- coding: utf-8 -*-
+# \file dfc_base.py
+# \brief TODO
+# \author Florent Guiotte <florent.guiotte@gmail.com>
+# \version 0.1
+# \date 27 août 2018
+#
+# TODO details
+
+import numpy as np
+
+import sys
+sys.path.append('..')
+import ld2dap
+
+def run(rasters, treshold=1e4):
+    # Parse parameters type
+    treshold = float(treshold)
+
+    # Pipelines
+    loader = ld2dap.LoadTIFF(rasters)
+    dfc_filter = ld2dap.Treshold(treshold)
+    dfc_filter.input = loader
+    out_vectors = ld2dap.RawOutput()
+    out_vectors.input = dfc_filter
+
+    # Compute vectors
+    out_vectors.run()
+
+    return out_vectors.data
+
+def version():
+    return 'v0.0'
diff --git a/descriptors/dfc_daps.py b/descriptors/dfc_daps.py
new file mode 100644
index 0000000..12c59fb
--- /dev/null
+++ b/descriptors/dfc_daps.py
@@ -0,0 +1,41 @@
+#!/usr/bin/python
+# -*- coding: utf-8 -*-
+# \file dfc_daps.py
+# \brief TODO
+# \author Florent Guiotte <florent.guiotte@gmail.com>
+# \version 0.1
+# \date 27 août 2018
+#
+# TODO details
+
+import numpy as np
+
+import sys
+sys.path.append('..')
+import ld2dap
+
+def run(rasters, treshold=1e4, areas=None, sd=None, moi=None):
+    # Parse parameters type
+    treshold = float(treshold)
+    areas = None if areas is None else np.array(areas).astype(np.float).astype(np.int)
+    sd = None if sd is None else np.array(sd).astype(np.float)
+    moi = None if moi is None else np.array(moi).astype(np.float)
+
+    # Pipelines
+    loader = ld2dap.LoadTIFF(rasters)
+    dfc_filter = ld2dap.Treshold(treshold)
+    dfc_filter.input = loader
+    aps = ld2dap.AttributeProfiles(area=areas, sd=sd, moi=moi)
+    aps.input = dfc_filter
+    differential = ld2dap.Differential()
+    differential.input = aps
+    out_vectors = ld2dap.RawOutput()
+    out_vectors.input = differential
+
+    # Compute vectors
+    out_vectors.run()
+
+    return out_vectors.data
+
+def version():
+    return 'v0.0'
diff --git a/descriptors/dfc_dsdaps.py b/descriptors/dfc_dsdaps.py
new file mode 100644
index 0000000..dca083d
--- /dev/null
+++ b/descriptors/dfc_dsdaps.py
@@ -0,0 +1,41 @@
+#!/usr/bin/python
+# -*- coding: utf-8 -*-
+# \file dfc_dsdaps.py
+# \brief TODO
+# \author Florent Guiotte <florent.guiotte@gmail.com>
+# \version 0.1
+# \date 28 août 2018
+#
+# TODO details
+
+import numpy as np
+
+import sys
+sys.path.append('..')
+import ld2dap
+
+def run(rasters, treshold=1e4, areas=None, sd=None, moi=None):
+    # Parse parameters type
+    treshold = float(treshold)
+    areas = None if areas is None else np.array(areas).astype(np.float).astype(np.int)
+    sd = None if sd is None else np.array(sd).astype(np.float)
+    moi = None if moi is None else np.array(moi).astype(np.float)
+
+    # Pipelines
+    loader = ld2dap.LoadTIFF(rasters)
+    dfc_filter = ld2dap.Treshold(treshold)
+    dfc_filter.input = loader
+    sdaps = ld2dap.SelfDualAttributeProfiles(area=areas, sd=sd, moi=moi)
+    sdaps.input = dfc_filter
+    differential = ld2dap.Differential()
+    differential.input = sdaps
+    out_vectors = ld2dap.RawOutput()
+    out_vectors.input = differential
+
+    # Compute vectors
+    out_vectors.run()
+
+    return out_vectors.data
+
+def version():
+    return 'v0.0'
diff --git a/descriptors/dfc_lfaps.py b/descriptors/dfc_lfaps.py
new file mode 100644
index 0000000..50853eb
--- /dev/null
+++ b/descriptors/dfc_lfaps.py
@@ -0,0 +1,57 @@
+#!/usr/bin/python
+# -*- coding: utf-8 -*-
+# \file dfc_lfaps.py
+# \brief TODO
+# \author Florent Guiotte <florent.guiotte@gmail.com>
+# \version 0.1
+# \date 27 août 2018
+#
+# TODO details
+
+import numpy as np
+
+import sys
+sys.path.append('..')
+import ld2dap
+
+# TODO: Add param percentile?
+
+dispatcher = {
+        'mean': np.mean, # Arithmetic mean
+        'median': np.median, # Median
+        'average': np.average, # Weighted average (=mean ?)
+        'std': np.std, # Standard deviation
+        'var': np.var, # Variance
+        'amax': np.amax, # Maximum
+        'amin': np.amin, # Minimum
+        'ptp': np.ptp, # Range of values (max - min)
+        }
+
+def run(rasters, treshold=1e4, areas=None, sd=None, moi=None, features=['mean'], patch_size=3):
+    # Parse parameters type
+    treshold = float(treshold)
+    areas = None if areas is None else np.array(areas).astype(np.float).astype(np.int)
+    sd = None if sd is None else np.array(sd).astype(np.float)
+    moi = None if moi is None else np.array(moi).astype(np.float)
+    patch_size = int(patch_size)
+    features = [dispatcher[x] for x in features]
+
+    # Pipelines
+    loader = ld2dap.LoadTIFF(rasters)
+    dfc_filter = ld2dap.Treshold(treshold)
+    dfc_filter.input = loader
+    aps = ld2dap.AttributeProfiles(area=areas, sd=sd, moi=moi)
+    aps.input = dfc_filter
+    local_features = ld2dap.LocalFeatures(features, patch_size)
+    local_features.input = aps
+    out_vectors = ld2dap.RawOutput()
+    out_vectors.input = local_features
+
+    # Compute vectors
+    out_vectors.run()
+
+    return out_vectors.data
+
+def version():
+    return 'v0.0'
+
diff --git a/descriptors/dfc_lfsdaps.py b/descriptors/dfc_lfsdaps.py
new file mode 100644
index 0000000..ab1556b
--- /dev/null
+++ b/descriptors/dfc_lfsdaps.py
@@ -0,0 +1,57 @@
+#!/usr/bin/python
+# -*- coding: utf-8 -*-
+# \file dfc_lfsdaps.py
+# \brief TODO
+# \author Florent Guiotte <florent.guiotte@gmail.com>
+# \version 0.1
+# \date 28 août 2018
+#
+# TODO details
+
+import numpy as np
+
+import sys
+sys.path.append('..')
+import ld2dap
+
+# TODO: Add param percentile?
+
+dispatcher = {
+        'mean': np.mean, # Arithmetic mean
+        'median': np.median, # Median
+        'average': np.average, # Weighted average (=mean ?)
+        'std': np.std, # Standard deviation
+        'var': np.var, # Variance
+        'amax': np.amax, # Maximum
+        'amin': np.amin, # Minimum
+        'ptp': np.ptp, # Range of values (max - min)
+        }
+
+def run(rasters, treshold=1e4, areas=None, sd=None, moi=None, features=['mean'], patch_size=3):
+    # Parse parameters type
+    treshold = float(treshold)
+    areas = None if areas is None else np.array(areas).astype(np.float).astype(np.int)
+    sd = None if sd is None else np.array(sd).astype(np.float)
+    moi = None if moi is None else np.array(moi).astype(np.float)
+    patch_size = int(patch_size)
+    features = [dispatcher[x] for x in features]
+
+    # Pipelines
+    loader = ld2dap.LoadTIFF(rasters)
+    dfc_filter = ld2dap.Treshold(treshold)
+    dfc_filter.input = loader
+    sdaps = ld2dap.SelfDualAttributeProfiles(area=areas, sd=sd, moi=moi)
+    sdaps.input = dfc_filter
+    local_features = ld2dap.LocalFeatures(features, patch_size)
+    local_features.input = sdaps
+    out_vectors = ld2dap.RawOutput()
+    out_vectors.input = local_features
+
+    # Compute vectors
+    out_vectors.run()
+
+    return out_vectors.data
+
+def version():
+    return 'v0.0'
+
diff --git a/descriptors/dfc_sdaps.py b/descriptors/dfc_sdaps.py
new file mode 100644
index 0000000..ab67dba
--- /dev/null
+++ b/descriptors/dfc_sdaps.py
@@ -0,0 +1,40 @@
+#!/usr/bin/python
+# -*- coding: utf-8 -*-
+# \file dfc_sdaps.py
+# \brief TODO
+# \author Florent Guiotte <florent.guiotte@gmail.com>
+# \version 0.1
+# \date 27 août 2018
+#
+# TODO details
+
+import numpy as np
+
+import sys
+sys.path.append('..')
+import ld2dap
+
+def run(rasters, treshold=1e4, areas=None, sd=None, moi=None):
+    # Parse parameters type
+    treshold = float(treshold)
+    areas = None if areas is None else np.array(areas).astype(np.float).astype(np.int)
+    sd = None if sd is None else np.array(sd).astype(np.float)
+    moi = None if moi is None else np.array(moi).astype(np.float)
+
+    # Pipelines
+    loader = ld2dap.LoadTIFF(rasters)
+    dfc_filter = ld2dap.Treshold(treshold)
+    dfc_filter.input = loader
+    sdaps = ld2dap.SelfDualAttributeProfiles(area=areas, sd=sd, moi=moi)
+    sdaps.input = dfc_filter
+    out_vectors = ld2dap.RawOutput()
+    out_vectors.input = sdaps
+
+    # Compute vectors
+    out_vectors.run()
+
+    return out_vectors.data
+
+def version():
+    return 'v0.0'
+
diff --git a/logger.py b/logger.py
new file mode 100644
index 0000000..14441fa
--- /dev/null
+++ b/logger.py
@@ -0,0 +1,51 @@
+#!/usr/bin/python
+# -*- coding: utf-8 -*-
+# \file %filename%.py
+# \brief TODO
+# \author Florent Guiotte <florent.guiotte@gmail.com>
+# \version 0.1
+# \date 24 avril 2018
+#
+# from https://fangpenlin.com/posts/2012/08/26/good-logging-practice-in-python/
+
+import os
+import logging.config
+from pathlib import Path
+
+import yaml
+
+def setup_logging(
+    default_path='logging.yaml',
+    default_level=logging.WARN,
+    env_key='LOG_CFG'
+):
+    """Setup logging configuration
+
+    """
+    path = default_path
+    value = os.getenv(env_key, None)
+    if value:
+        path = value
+    if os.path.exists(path):
+        with open(path, 'rt') as f:
+            config = yaml.safe_load(f.read())
+        makedirs(config)
+        logging.config.dictConfig(config)
+    else:
+        logging.basicConfig(level=default_level)
+
+
+def makedirs(dic):
+    files = finddirs(dic)
+    for f in files:
+        d = Path(*f.parts[:-1])
+        d.mkdir(parents=True, exist_ok=True)
+
+def finddirs(dic, key='filename'):
+    r = list()
+    value = dic.get(key)
+    if value : r.append(Path(value))
+    for k, v in dic.items():
+        if isinstance(v, dict):
+            r.extend(finddirs(v))
+    return r
diff --git a/logging.yaml b/logging.yaml
new file mode 100644
index 0000000..b18e6ed
--- /dev/null
+++ b/logging.yaml
@@ -0,0 +1,40 @@
+version: 1
+disable_existing_loggers: False
+formatters:
+    simple:
+        format: "%(asctime)s %(levelname)s %(name)s: %(message)s"
+
+handlers:
+    console:
+        class: logging.StreamHandler
+        level: INFO
+        formatter: simple
+        stream: ext://sys.stdout
+
+    info_file_handler:
+        class: logging.handlers.RotatingFileHandler
+        level: INFO
+        formatter: simple
+        filename: Logs/info.log
+        maxBytes: 10485760 # 10MB
+        backupCount: 20
+        encoding: utf8
+
+    error_file_handler:
+        class: logging.handlers.RotatingFileHandler
+        level: ERROR
+        formatter: simple
+        filename: Logs/errors.log
+        maxBytes: 10485760 # 10MB
+        backupCount: 20
+        encoding: utf8
+
+loggers:
+    my_module:
+        level: DEBUG
+        handlers: [console]
+        propagate: no
+
+root:
+    level: DEBUG
+    handlers: [console, info_file_handler, error_file_handler]
diff --git a/protocols/__init__.py b/protocols/__init__.py
new file mode 100644
index 0000000..e17274e
--- /dev/null
+++ b/protocols/__init__.py
@@ -0,0 +1,9 @@
+#!/usr/bin/python
+# -*- coding: utf-8 -*-
+# \file __init__.py
+# \brief TODO
+# \author Florent Guiotte <florent.guiotte@gmail.com>
+# \version 0.1
+# \date 09 sept. 2018
+#
+# TODO details
diff --git a/protocols/jurse.py b/protocols/jurse.py
new file mode 100644
index 0000000..1f41485
--- /dev/null
+++ b/protocols/jurse.py
@@ -0,0 +1,134 @@
+#!/usr/bin/python
+# -*- coding: utf-8 -*-
+# \file jurse.py
+# \brief TODO
+# \author Florent Guiotte <florent.guiotte@gmail.com>
+# \version 0.1
+# \date 07 sept. 2018
+#
+# TODO details
+
+import hashlib
+import importlib
+from collections import OrderedDict
+import numpy as np
+import pandas as pd
+from sklearn import metrics
+import triskele
+from .protocol import Protocol, TestError
+
+
+class Jurse(Protocol):
+    """First JURSE test protocol for LiDAR classification with 2D maps.
+
+    This first protocol compute attribute profiles on the whole scene then
+    split in train and test for a random forest classifier.
+
+    """
+
+    def __init__(self, expe):
+        super().__init__(expe, self.__class__.__name__)
+
+    def _get_hashes(self):
+        hashes = OrderedDict()
+        glob = hashlib.sha1()
+
+        for k in ['ground_truth', 'descriptors_script',
+                  'cross_validation', 'classifier']:
+            val = str(self._expe[k]).encode('utf-8')
+            hashes[k] = hashlib.sha1(val).hexdigest()
+            glob.update(val)
+        hashes['global'] = glob.hexdigest()
+
+        return hashes
+
+    def _run(self):
+        self._log.info('Compute descriptors')
+        try:
+            descriptors = self._compute_descriptors()
+        except Exception:
+            raise TestError('Error occured during description')
+        self._time('description')
+
+        self._log.info('Classify data')
+        try:
+            classification = self._compute_classification(descriptors)
+        except Exception:
+            raise TestError('Error occured during classification')
+        self._time('classification')
+
+        self._log.info('Run metrics')
+        metrics = self._run_metrics(classification, descriptors)
+        self._time('metrics')
+
+        cmap = str(self._results_base_name) + '.tif'
+        self._log.info('Saving classification map {}'.format(cmap))
+        triskele.write(cmap, classification)
+
+        results = OrderedDict()
+        results['classification'] = cmap
+        results['metrics'] = metrics
+        self._results = results
+
+    def _compute_descriptors(self):
+        script = self._expe['descriptors_script']
+
+        desc = importlib.import_module(script['name'])
+        att = desc.run(**script['parameters'])
+
+        return att
+
+    def _compute_classification(self, descriptors):
+        # Ground truth
+        gt = self._get_ground_truth()
+
+        # CrossVal and ML
+        cv = self._expe['cross_validation']
+        cl = self._expe['classifier']
+
+        cross_val = getattr(importlib.import_module(cv['package']), cv['name'])
+        classifier = getattr(importlib.import_module(cl['package']), cl['name'])
+
+        prediction = np.zeros_like(gt, dtype=np.uint8)
+
+        for xt, xv, yt, yv, ti in cross_val(gt, descriptors, **cv['parameters']):
+            rfc = classifier(**cl['parameters'])
+            rfc.fit(xt, yt)
+
+            ypred = rfc.predict(xv)
+
+            prediction[ti] = ypred
+
+        return prediction
+
+    def _get_ground_truth(self):
+        gt_expe = self._expe['ground_truth']
+        gt = triskele.read(gt_expe['raster'])
+
+        # Meta labeling
+        idx_map = np.arange(gt.max() + 1)
+
+        if 'meta_labels' in gt_expe:
+            meta_idx = pd.read_csv(gt_expe['meta_labels'])
+            idx = np.array(meta_idx['index'])
+            midx = np.array(meta_idx['metaclass_index'])
+            idx_map[idx] = midx
+
+        return idx_map[gt]
+
+    def _get_results(self):
+        return self._results
+
+    def _run_metrics(self, classification, descriptors):
+        gt = self._get_ground_truth()
+
+        f = np.nonzero(classification)
+        pred = classification[f].ravel()
+        gt = gt[f].ravel()
+
+        results = OrderedDict()
+        results['dimensions'] = descriptors.shape[-1]
+        results['overall_accuracy'] = float(metrics.accuracy_score(gt, pred))
+        results['cohen_kappa'] = float(metrics.cohen_kappa_score(gt, pred))
+
+        return results
diff --git a/protocols/protocol.py b/protocols/protocol.py
new file mode 100644
index 0000000..b53efc8
--- /dev/null
+++ b/protocols/protocol.py
@@ -0,0 +1,67 @@
+#!/usr/bin/python
+# -*- coding: utf-8 -*-
+# \file protocol.py
+# \brief TODO
+# \author Florent Guiotte <florent.guiotte@gmail.com>
+# \version 0.1
+# \date 07 sept. 2018
+#
+# TODO details
+
+import logging
+import time
+from collections import OrderedDict
+
+
+class Protocol:
+    def __init__(self, expe, name=None):
+        self._log = logging.getLogger(name)
+        self._expe = expe
+        self._name = name
+        self._times = OrderedDict()
+        self._results_base_name = None
+        self._log.debug('expe loaded: {}'.format(self._expe))
+
+    def get_hashes(self):
+        self._log.info('Computing hashes')
+        return(self._get_hashes())
+
+    def set_results_base_name(self, base_name):
+        self._results_base_name = base_name
+
+    def run(self):
+        self._pt = time.process_time()
+        self._run()
+
+    def get_results(self):
+        return self._get_results()
+
+    def get_process_time(self):
+        return self._times
+
+    def _time(self, process):
+        self._times[process] = time.process_time() - self._pt
+        self._pt = time.process_time()
+
+    def _get_hashes(self):
+        self._log.warning('Protocol did not override _get_hashes()')
+        hashes = OrderedDict()
+        hashes['global'] = 'Protocol did not override _get_hashes()'
+        return hashes
+
+    def _run(self):
+        self._log.error('Protocol did not override _run()')
+        raise NotImplementedError('Protocol {} did not override _run()'.format(self))
+
+    def _get_results(self):
+        self._log.warning('Protocol did not override _get_results()')
+        results = OrderedDict()
+        results['global'] = 'Protocol did not override _get_results()'
+        return results
+
+    def __str__(self):
+        return('{}'.format(self._name))
+
+
+class TestError(Exception):
+    pass
diff --git a/result_mockup.yml b/result_mockup.yml
new file mode 100644
index 0000000..f2e8a6d
--- /dev/null
+++ b/result_mockup.yml
@@ -0,0 +1,65 @@
+name: Première expérience
+date: 9 juillet 2018
+priority: 1
+detail: |
+    Maquette pour la création du supervisor de minigrida.  Par rapport à la
+    version legacy du projet LD2DAPs, le choix du protocole de test est
+    générique. Il faut ajouter le chargement dynamique du protocole puis
+    réusiner le fonctionnement du supervisor pour respecter l'esprit universel
+    de minigrida.
+protocol: 
+    name: Jurse
+    package: protocols.jurse
+expe:
+  ground_truth: 
+    raster: ./Data/ground_truth/2018_IEEE_GRSS_DFC_GT_TR.tif
+    meta_labels: ./Data/ground_truth/jurse_meta_idx.csv
+  descriptors_script:
+    name: descriptors.dfc_aps
+    parameters:
+      areas:
+      - 100
+      - 1000
+      moi:
+      - 0.5
+      - 0.9
+      rasters:
+      - ./Data/dfc_rasters/DEM+B_C123/UH17_GEM051_TR.tif
+      - ./Data/dfc_rasters/DEM_C123_3msr/UH17_GEG051_TR.tif
+      treshold: 1e4
+  cross_validation:
+    name: APsCVG
+    package: CVGenerators
+    parameters:
+      n_test: 2
+  classifier:
+    name: RandomForestClassifier
+    package: sklearn.ensemble
+    parameters:
+      min_samples_leaf: 10
+      n_estimators: 50
+      n_jobs: -1
+      random_state: 0
+hashes:
+  ground_truth: 2c5ecaddcb8c4a1c8863bc65e7440de4a1b4962c
+  descriptors_script: cfdcc84d9d9c47177930257f286d850db446812b
+  cross_validation: 4a61b34fda812fe717890b25d75430023335a7a6
+  classifier: 40e6741ef8cc4b4fbe188b8ca0563eb5195b88ad
+  global: b8219fab322bf11ec1aac14a1f51466dd94ddbdd
+report:
+  supervisor: thecomedian
+  start_date: Le 27/07/2018 à 16:28:52
+  end_date: Le 27/07/2018 à 16:29:54
+  ressources:
+    ram: 42 Go
+    process_time:
+      description: 0.6744262149950373
+      classification: 168.82905034400028
+      metrics: 1.1557443889978458
+results:
+  classification: ./Enrichment/Results/test_b8219f.tif
+  metrics:
+    dimensions: 42
+    overall_accuracy: 0.5550408093111998
+    cohen_kappa: 0.41714275852261407
+
diff --git a/supervisor.py b/supervisor.py
new file mode 100644
index 0000000..bada05e
--- /dev/null
+++ b/supervisor.py
@@ -0,0 +1,235 @@
+#!/usr/bin/python
+# -*- coding: utf-8 -*-
+# \file supervisor.py
+# \brief TODO
+# \author Florent Guiotte <florent.guiotte@gmail.com>
+# \version 0.1
+# \date 07 sept. 2018
+#
+# TODO details
+
+import yaml
+import importlib
+import hashlib
+from collections import OrderedDict
+import time
+import os
+import datetime
+from pathlib import Path
+from operator import itemgetter
+import traceback
+import logging
+import logger
+from protocols.protocol import TestError
+
+ENRICHMENT_DIR = Path('./Enrichment/')
+TEST_DIR    = ENRICHMENT_DIR / 'Tests'
+STAGING_DIR = ENRICHMENT_DIR / 'Staging'
+RESULT_DIR  = ENRICHMENT_DIR / 'Results'
+FAILED_DIR  = ENRICHMENT_DIR / 'Failed'
+
+log = logging.getLogger('Supervisor [{}]'.format(os.uname()[1]))
+
+def setup_yaml():
+    """ Keep yaml ordered, newline string
+    from https://stackoverflow.com/a/8661021
+    """
+    represent_dict_order = lambda self, data: self.represent_mapping('tag:yaml.org,2002:map', data.items())
+    yaml.add_representer(OrderedDict, represent_dict_order)
+
+    """ https://stackoverflow.com/a/24291536 """
+    yaml.Dumper.org_represent_str = yaml.Dumper.represent_str
+    yaml.add_representer(str, repr_str, Dumper=yaml.Dumper)
+
+
+def repr_str(dumper, data):
+    if '\n' in data:
+        return dumper.represent_scalar(u'tag:yaml.org,2002:str',
+                                       data, style='|')
+    return dumper.org_represent_str(data)
+
+
+def update_queue():
+    tmp_queue = list()
+    for child in TEST_DIR.iterdir():
+        if child.is_file() and child.suffix == '.yml':
+            tmp_queue.append({'expe_file': ExpePath(child),
+                              'priority': get_priority(child)})
+
+    queue = sorted(tmp_queue, key=itemgetter('priority'))
+    return queue
+
+
+def get_priority(yml_file):
+    with open(yml_file) as f:
+        expe = OrderedDict(yaml.safe_load(f))
+    return expe['priority']
+
+
+def run(expe_file):
+    start_time = time.time()
+    log.info('Run test {}'.format(expe_file))
+    test = expe_file.read()
+
+    ### Stage experience
+    expe_file.stage(test)
+
+    ### Load protocol
+    try:
+        #protocol = getattr(importlib.import_module(test['protocol']['package']), test['protocol']['name'])
+        protocol_module = importlib.import_module(test['protocol']['package'])
+        importlib.reload(protocol_module)
+        protocol = getattr(protocol_module, test['protocol']['name'])
+        experience = protocol(test['expe'])
+    except Exception as e:
+        err = 'Could not load protocol from test {}'.format(expe_file)
+        log.warning(err)
+        expe_file.error(test, 'loading protocol', e)
+        raise TestError(err)
+    log.info('{} test protocol loaded'.format(experience))
+
+    ### Get hashes
+    test['hashes'] = experience.get_hashes()
+    test['report'] = create_report(experience, start_time)
+
+    ### Stage experience
+    expe_file.stage(test)
+
+    experience.set_results_base_name(expe_file.get_result_path())
+
+    ### Run test
+    try:
+        experience.run()
+    except Exception as e:
+        err = 'Experience error'
+        log.warning(err)
+        expe_file.error(test, 'testing', e)
+        raise TestError(err) 
+
+    end_time = time.time()
+
+    ### Get complete report
+    test['report'] = create_report(experience, start_time, end_time)
+
+    ### Get results
+    test['results'] = experience.get_results()
+
+    ### Write experience
+    expe_file.result(test)
+
+    ### End of test
+    log.info('Test complete')
+
+def create_report(experience, stime=None, etime=None):
+    expe_report = OrderedDict()
+    expe_report['supervisor'] = os.uname()[1]
+
+    # Dates
+    for datek, timev in zip(('start_date', 'end_date'), (stime, etime)):
+        expe_report[datek] = datetime.datetime.fromtimestamp(timev).strftime('Le %d/%m/%Y à %H:%M:%S') if timev is not None else None
+
+    # Ressources
+    ressources = OrderedDict()
+    ressources['ram'] = None
+    ressources['proccess_time'] = experience.get_process_time()
+    expe_report['ressources'] = ressources
+
+    return expe_report
+
+class ExpePath:
+    """Utility wrapper for expe files.
+
+    Extend pathlib Path with staging, result and errors function to move the
+    test report through the Enrichment center.
+
+    """
+    def __init__(self, path, hash_length=6):
+        self._actual = Path(path)
+        self._base_name = self._actual.stem
+        self._hash_length = hash_length
+        self._hash = None
+
+    def __str__(self):
+        return self._get_complete_name()
+
+    def read(self):
+        with open(self._actual) as f:
+            return OrderedDict(yaml.safe_load(f))
+
+    def _get_hash_name(self):
+        return '{}{}'.format(self._base_name,
+                '_' + self._hash[:self._hash_length] if self._hash is not None else '')
+
+    def _get_complete_name(self):
+        return self._get_hash_name() + '.yml'
+
+    def stage(self, expe):
+        log.info('Staging {}'.format(self._base_name))
+        self._check_hash(expe)
+        self._write(STAGING_DIR, expe)
+
+    def result(self, expe):
+        log.info('Write results for test {}'.format(self._base_name))
+        self._check_hash(expe)
+        self._write(RESULT_DIR, expe)
+
+    def error(self, expe, when='', e=Exception):
+        error = OrderedDict()
+        error['when'] = when
+        error['what'] = str(e)
+        error['where'] = traceback.format_exc()
+        expe['error'] = error
+        self._write(FAILED_DIR, expe)
+
+    def get_result_path(self):
+        return Path(RESULT_DIR) / self._get_hash_name()
+
+    def _check_hash(self, expe):
+        if self._hash is None:
+            if 'hashes' in expe:
+                self._hash = expe['hashes']['global']
+
+    def _write(self, path, expe):
+        new_path = Path(path) / self._get_complete_name()
+        with open(new_path, 'w') as of:
+            yaml.dump(expe, of, 
+                      default_flow_style=False, 
+                      encoding=None, 
+                      allow_unicode=True)
+        self._actual.unlink()
+        self._actual = new_path
+    
+
+def watch_folder():
+    log.info('Waiting for test')
+    while not list(TEST_DIR.glob('*.yml')):
+        time.sleep(3)
+
+def main():
+    while(True):
+        try:
+            queue = update_queue()
+        except Exception:
+            log.error('Critical exception while updating work queue')
+            log.error(traceback.format_exc())
+            log.warning('Resuming')
+            continue
+        if not queue:
+            watch_folder()
+            continue
+        try:
+            run(queue.pop()['expe_file'])
+        except TestError:
+            log.warning('Test failed, error logged. Resuming')
+        except Exception:
+            log.error('Critical exception while running test. Resuming')
+            log.error(traceback.format_exc())
+            log.warning('Resuming')
+            continue
+    
+if __name__ == '__main__':
+    logger.setup_logging()
+    log.info('Starting supervisor')
+
+    setup_yaml()
+    main()
diff --git a/test_mockup.yml b/test_mockup.yml
new file mode 100644
index 0000000..503caee
--- /dev/null
+++ b/test_mockup.yml
@@ -0,0 +1,42 @@
+name: Première expérience
+date: 9 juillet 2018
+priority: 1
+detail: |
+    Maquette pour la création du supervisor de minigrida.  Par rapport à la
+    version legacy du projet LD2DAPs, le choix du protocole de test est
+    générique. Il faut ajouter le chargement dynamique du protocole puis
+    réusiner le fonctionnement du supervisor pour respecter l'esprit universel
+    de minigrida.
+protocol: 
+    name: Jurse
+    package: protocols.jurse
+expe:
+  ground_truth: 
+    raster: ./Data/ground_truth/2018_IEEE_GRSS_DFC_GT_TR.tif
+    meta_labels: ./Data/ground_truth/jurse_meta_idx.csv
+  descriptors_script:
+    name: descriptors.dfc_aps
+    parameters:
+      areas:
+      - 100
+      - 1000
+      moi:
+      - 0.5
+      - 0.9
+      rasters:
+      - ./Data/dfc_rasters/DEM+B_C123/UH17_GEM051_TR.tif
+      - ./Data/dfc_rasters/DEM_C123_3msr/UH17_GEG051_TR.tif
+      treshold: 1e4
+  cross_validation:
+    name: APsCVG
+    package: CVGenerators
+    parameters:
+      n_test: 2
+  classifier:
+    name: RandomForestClassifier
+    package: sklearn.ensemble
+    parameters:
+      min_samples_leaf: 10
+      n_estimators: 10
+      n_jobs: -1
+      random_state: 0