diff --git a/dimos/core/resource.py b/dimos/core/resource.py
index 21cdec6322..ce3f735329 100644
--- a/dimos/core/resource.py
+++ b/dimos/core/resource.py
@@ -21,3 +21,25 @@ def start(self) -> None: ...
 
     @abstractmethod
     def stop(self) -> None: ...
+
+    def dispose(self) -> None:
+        """
+        Makes a Resource disposable
+        So you can do a
+
+        from reactivex.disposable import CompositeDisposable
+
+        disposables = CompositeDisposable()
+
+        transport1 = LCMTransport(...)
+        transport2 = LCMTransport(...)
+
+        disposables.add(transport1)
+        disposables.add(transport2)
+
+        ...
+
+        disposables.dispose()
+
+        """
+        self.stop()
diff --git a/dimos/perception/detection/detectors/test_bbox_detectors.py b/dimos/perception/detection/detectors/test_bbox_detectors.py
index 4047e35e7b..32a509061a 100644
--- a/dimos/perception/detection/detectors/test_bbox_detectors.py
+++ b/dimos/perception/detection/detectors/test_bbox_detectors.py
@@ -12,8 +12,12 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
+from dimos_lcm.foxglove_msgs.ImageAnnotations import ImageAnnotations
 import pytest
+from reactivex.disposable import CompositeDisposable
 
+from dimos.core import LCMTransport
+from dimos.msgs.sensor_msgs import Image
 from dimos.perception.detection.type import Detection2D, ImageDetections2D
 
 
@@ -24,9 +28,37 @@ def detector(request):
 
 
 @pytest.fixture(scope="session")
-def detections(detector, test_image):
+def get_topic_annotations():
+    disposables = CompositeDisposable()
+
+    def topic_annotations(suffix: str = "unnamed"):
+        annotations: LCMTransport[ImageAnnotations] = LCMTransport(
+            f"/annotations_{suffix}", ImageAnnotations
+        )
+        disposables.add(annotations)
+        return annotations
+
+    yield topic_annotations
+    disposables.dispose()
+
+
+@pytest.fixture(scope="session")
+def detections(detector, test_image, topic_image, get_topic_annotations):
     """Get ImageDetections2D from any detector."""
-    return detector.process_image(test_image)
+    topic_image.publish(test_image)
+    detections = detector.process_image(test_image)
+    annotations = detections.to_foxglove_annotations()
+    print("annotations:", annotations)
+    topic_annotations = get_topic_annotations(detector.__class__.__name__)
+    topic_annotations.publish(annotations)
+    return detections
+
+
+@pytest.fixture(scope="session")
+def topic_image():
+    image: LCMTransport[Image] = LCMTransport("/color_image", Image)
+    yield image
+    image.lcm.stop()
 
 
 def test_detection_basic(detections) -> None: