diff --git a/docs/group_assignments/group2/assets/assignment3/kerf-pieces.JPG b/docs/group_assignments/group2/assets/assignment3/kerf-pieces.JPG
new file mode 100644
index 0000000000000000000000000000000000000000..3ab4c539b74bf7f4b628ba81768e1edb1e267682
Binary files /dev/null and b/docs/group_assignments/group2/assets/assignment3/kerf-pieces.JPG differ
diff --git a/docs/group_assignments/group2/assets/assignment3/kerf-with-without.JPG b/docs/group_assignments/group2/assets/assignment3/kerf-with-without.JPG
new file mode 100644
index 0000000000000000000000000000000000000000..997a91cf7be92dc54a9de23ad3856b92ab5cfc7a
Binary files /dev/null and b/docs/group_assignments/group2/assets/assignment3/kerf-with-without.JPG differ
diff --git a/docs/group_assignments/group2/assignment3.md b/docs/group_assignments/group2/assignment3.md
index 94194d68727be2dacbb673ed37f93778563a900f..bbfcbe34124587069715a612f9ba577b2084c269 100644
--- a/docs/group_assignments/group2/assignment3.md
+++ b/docs/group_assignments/group2/assignment3.md
@@ -106,12 +106,14 @@ Then we cut pieces, where the kerf was taken in to account. We used the value fo
 kerf based on the measurements we made first in the group work. In these cases the pieces were well attached. It was even
 difficult to push the pieces together. The reason for this was that the material that we cut was thicker than the one we initially measured.
 
-Figure of our parts
+![kerf](./assets/assignment3/kerf-with-without.JPG)
+Figure of pieces with and without compensating for kerf
 
 We tested with the pieces, where the correct thickness was used in the design. For one pair, the kerf was not taken in to
 account in the other part, and it was loose. For the other pair the kerf was correctly taken in to account, and the connection
 was good.
 
+![kerf](./assets/assignment3/kerf-pieces.JPG)
 Figure of the ready made parts
 
 mkdocs formatting examples (remove in the end):