diff --git a/arxiv/utils_restreamer.py b/arxiv/utils_restreamer.py
index 0b88f77..74f0117 100644
--- a/arxiv/utils_restreamer.py
+++ b/arxiv/utils_restreamer.py
@@ -182,7 +182,7 @@ def write_to_rabbit_csv(final_data_frame):
         row = final_data_frame.iloc[num]  # this a dataframe that
         # corresponds to the num line of the big dataframe
         csv_string=final_data_frame[num:(num+1)].to_csv(path_or_buf=None,
-                                                     index=False,
+                                                        index=False,
                                                         header=False)
         # csv_string = StringIO()
         # row.to_csv(csv_string, header=False)
@@ -573,8 +573,7 @@ def main():
     logging.basicConfig(level=logging.INFO, format=LOG_FORMAT)
 
     validate_conf(gconf)
-    config_dict = read_or_init_restreamed_metadata(
-        METADATA_FILE)
+    config_dict = read_or_init_restreamed_metadata(METADATA_FILE)
     try:
         number_of_executions = gconf['numberOfDaysToReplay']
         for i in range(number_of_executions):
diff --git a/dsio/dashboard/kibana.py b/dsio/dashboard/kibana.py
index 15c252b..cc0e6f8 100644
--- a/dsio/dashboard/kibana.py
+++ b/dsio/dashboard/kibana.py
@@ -4,7 +4,7 @@
 from kibana_dashboard_api import VisualizationsManager, DashboardsManager
 
 
-def generate_dashboard(es_conn, sensor_names, df_scored, index_name):
+def generate_dashboard(es_conn, sensor_names, df_scored, index_name, timefield='time'):
     dashboards = DashboardsManager(es_conn)
     dashboard = Dashboard()
     dashboard.id = "%s-dashboard" % index_name
@@ -67,12 +67,31 @@ def generate_dashboard(es_conn, sensor_names, df_scored, index_name):
                 },
                 "aggs": [
                     {
+                        "id": "1",
+                        "type": "avg",
+                        "schema":"metric",
+                        "params": {
+                            "field": sensor,
+                            "customLabel": sensor.replace('_', ' ')
+                        }
+                    }, {
                         "id": "2",
                         "type": "max",
                         "schema":"radius",
                         "params": {
                             "field":"SCORE_%s" % sensor
                         }
+                    }, {
+                        "id": "3",
+                        "type": "date_histogram",
+                        "schema": "segment",
+                        "params":{
+                            "field": timefield,
+                            "interval": "custom",
+                            "customInterval": "5s",
+                            "min_doc_count": 1,
+                            "extended_bounds": {}
+                        }
                     }
                 ],
                 "listeners": {}
@@ -96,10 +115,30 @@ def generate_dashboard(es_conn, sensor_names, df_scored, index_name):
         ret = dashboard.add_visualization(viz)
         i += 1
 
+    # Create the index if it does not exist
+    if not es_conn.indices.exists(index_name):
+        index_properties = {"time" : {"type": "date"}}
+        body = {"mappings": {index_name: {"properties": index_properties}}}
+        es_conn.indices.create(index=index_name, body=body)
+
     try:
         ret = dashboards.add(dashboard)
     except elasticsearch.exceptions.ConflictError:
         ret = dashboards.update(dashboard)
         #pass # Dashboard already exists, let's not overwrite it
 
+    es_conn.index(index='.kibana', doc_type="index-pattern", id=index_name,
+                  body={"title": index_name, "timeFieldName": "time"})
+
+    kibana_config = es_conn.search(index='.kibana',
+                                   sort={'_uid': {'order': 'desc'}},
+                                   doc_type='config')
+    try:
+        kibana_id = kibana_config['hits']['hits'][0]['_id']
+    except:
+        raise # TODO
+
+    es_conn.update(index='.kibana', doc_type='config', id=kibana_id,
+                   body={"doc": {"defaultIndex" : index_name}})
+
     return ret
diff --git a/dsio/restream/elastic.py b/dsio/restream/elastic.py
index e4f660d..3fcd0d0 100644
--- a/dsio/restream/elastic.py
+++ b/dsio/restream/elastic.py
@@ -1,8 +1,6 @@
 """
 Elasticsearch batch re-streamer
 """
-from __future__ import print_function
-
 import time
 import datetime
 
@@ -11,7 +9,7 @@
 from elasticsearch import helpers
 
 
-def batchRedater(X, timefield, hz = 10):
+def batchRedater(X, timefield, hz=10):
     # send 10 datapoints a second
     now = np.int(np.round(time.time()))
     X[timefield] = (now*1000 + X.index._data*hz)
@@ -36,14 +34,18 @@ def df2es(Y, index_name, es, index_properties=None, recreate=True,
             print('Deleting existing index {}'.format(index_name))
         except:
             pass
+
         print('Creating index {}'.format(index_name))
-        es.indices.create(index_name, body = body)
+        es.indices.create(index_name, body=body)
 
     # Formatting the batch to upload as a tuple of dictionnaries
     list_tmp = tuple(Y.fillna(0).T.to_dict().values())
+
     # Exporting to ES
     out = helpers.bulk(es, list_tmp)
 
+    return out
+
 
 # X = features
 def elasticsearch_batch_restreamer(X, timefield, es, index_name,
@@ -58,7 +60,7 @@ def elasticsearch_batch_restreamer(X, timefield, es, index_name,
         everyX = 200
 
     virtual_time = np.min(X[timefield])
-    recreate = True
+    recreate = False
     while virtual_time < np.max(X[timefield]):
         start_time = virtual_time
         virtual_time += everyX*1000
diff --git a/setup.py b/setup.py
index f4d34f7..7e7e8ff 100644
--- a/setup.py
+++ b/setup.py
@@ -38,7 +38,7 @@
         'Development Status :: 2 - Pre-Alpha',
         'Intended Audience :: Developers',
         'Natural Language :: English',
-        'Programming Language :: Python :: 2.7'
+        'Programming Language :: Python :: 3.5'
     ],
     entry_points={
         'console_scripts': ['dsio=dsio.main:main'],