diff --git a/src/constantinople_lab_to_nwb/general_interfaces/bpodbehaviorinterface.py b/src/constantinople_lab_to_nwb/general_interfaces/bpodbehaviorinterface.py
index 21bd738..f78649a 100644
--- a/src/constantinople_lab_to_nwb/general_interfaces/bpodbehaviorinterface.py
+++ b/src/constantinople_lab_to_nwb/general_interfaces/bpodbehaviorinterface.py
@@ -17,10 +17,11 @@
     TaskArgumentsTable,
 )
 from neuroconv import BaseDataInterface
-from neuroconv.utils import DeepDict
+from neuroconv.utils import DeepDict, get_base_schema, get_schema_from_hdmf_class
 
 from ndx_structured_behavior.utils import loadmat
 from pynwb import NWBFile
+from pynwb.device import Device
 
 
 class BpodBehaviorInterface(BaseDataInterface):
@@ -48,8 +49,39 @@ def __init__(
         self.file_path = file_path
         self._bpod_struct = self._read_file()
         self._block_name_mapping = {1: "Mixed", 2: "High", 3: "Low"}
+        self._trial_start_times = None
+        self._trial_stop_times = None
         super().__init__(file_path=file_path, verbose=verbose)
 
+    def get_metadata_schema(self) -> dict:
+        metadata_schema = super().get_metadata_schema()
+        metadata_schema["properties"]["Behavior"] = get_base_schema(tag="Behavior")
+        device_schema = get_schema_from_hdmf_class(Device)
+        metadata_schema["properties"]["Behavior"].update(
+            required=[
+                "Device",
+                "StateTypesTable",
+                "StatesTable",
+                "ActionTypesTable",
+                "ActionsTable",
+                "EventTypesTable",
+                "EventsTable",
+                "TrialsTable",
+            ],
+            properties=dict(
+                Device=device_schema,
+                StateTypesTable=dict(type="object", properties=dict()),
+                StatesTable=dict(type="object", properties=dict()),
+                ActionTypesTable=dict(type="object", properties=dict()),
+                ActionsTable=dict(type="object", properties=dict()),
+                EventTypesTable=dict(type="object", properties=dict()),
+                EventsTable=dict(type="object", properties=dict()),
+                TrialsTable=dict(type="object", properties=dict()),
+                TaskArgumentsTable=dict(type="object", properties=dict()),
+            ),
+        )
+        return metadata_schema
+
     def get_metadata(self) -> DeepDict:
         metadata = super().get_metadata()
 
@@ -81,6 +113,7 @@ def get_metadata(self) -> DeepDict:
             EventTypesTable=dict(description="Contains the name of the events in the task."),
             EventsTable=dict(description="Contains the onset times of events in the task."),
             TrialsTable=dict(description="Contains the start and end times of each trial in the task."),
+            TaskArgumentsTable=dict(description="Contains the task arguments for the task."),
         )
 
         task_arguments = dict(
@@ -301,7 +334,17 @@ def _read_file(self) -> dict:
         return mat_file[self.default_struct_name]
 
     def get_trial_times(self) -> (List[float], List[float]):
-        return self._bpod_struct["TrialStartTimestamp"], self._bpod_struct["TrialEndTimestamp"]
+        trial_start_times = (
+            self._trial_start_times if self._trial_start_times is not None else self._bpod_struct["TrialStartTimestamp"]
+        )
+        trial_stop_times = (
+            self._trial_stop_times if self._trial_stop_times is not None else self._bpod_struct["TrialEndTimestamp"]
+        )
+        return trial_start_times, trial_stop_times
+
+    def set_aligned_trial_times(self, trial_start_times: List[float], trial_stop_times: List[float]) -> None:
+        self._trial_start_times = trial_start_times
+        self._trial_stop_times = trial_stop_times
 
     def create_states_table(
         self, metadata: dict, trial_start_times: List[float]
@@ -519,7 +562,6 @@ def add_trials(self, nwbfile: NWBFile, metadata: dict) -> None:
             events_table=events_table,
             actions_table=actions_table,
         )
-        trial_stop_times = trial_start_times[1:] + [np.nan]
         for start, stop in zip(trial_start_times, trial_stop_times):
             states_table_df = states_table[:]
             states_index_mask = (states_table_df["start_time"] >= start) & (states_table_df["stop_time"] < stop)