diff --git a/BPV/qrs/presentation.typ b/BPV/qrs/presentation.typ
index 52c707b..c84f94f 100644
--- a/BPV/qrs/presentation.typ
+++ b/BPV/qrs/presentation.typ
@@ -20,6 +20,7 @@
 )
 
 #slide(title: "Introduction")[
+//add images with wein diagram from proposal presentation
 ]
 
 #slide(title: "Overview")[
@@ -32,9 +33,20 @@
 ]
 
 #slide(title: "BPV Model")[
+#image("images/training.svg", height: 100%)
+// fix font in inkscape, use STIX TWO Text everywhere
+// Remove one model to only present single model mode
 ]
 
 #slide(title: "Multi-Model Support")[
+// Reuse the training image but with both model
+// underline that the threshold seems similar bu they refer to possibly widely different traces.
+// one new trace does not have a single position on the bottom graph, it has one (possibly 
+// different) per model
+]
+
+#slide(title: "Multi-Model Support")[
+// Explain decision, trace needs to be validated by at least one model to pass
 ]
 
 #focus-slide()[
@@ -42,27 +54,30 @@ Test Cases
 ]
 
 #slide(title: "Test Case 1")[
-
+// table with devices and modifications/attacks
 ]
 
 #slide(title: "Test Case 1 - Results")[
+// table with results
 ]
 
 
 #slide(title: "Test Case 2")[
-
+// table with modifications to the drone
 ]
 
 #slide(title: "Test Case 2 - Results")[
 ]
 
-
+// if no results, present the setup and say why this is interesting and versatile
 #slide(title: "Test Case 3")[
 ]
 
 #slide(title: "Test Case 3 - Results")[
 ]
 
+// possibly a slide about anomaly generation, maybe in futur work
+
 #slide(title: "Future Work")[
 ]