diff --git a/src/local_search.py b/src/local_search.py
new file mode 100644
index 0000000..915d14f
--- /dev/null
+++ b/src/local_search.py
@@ -0,0 +1,69 @@
+import os
+if 'AI' in os.getcwd():
+    from src.utils import *
+else:
+    from AI2019.src.utils import *
+
+
+
+class TwoOpt:
+
+    @staticmethod
+    def step2opt(solution, matrix_dist, distance):
+        """
+        One step of 2opt, one double loop and return first improved sequence
+        @param tsp_sequence:
+        @param matrix_dist:
+        @param distance:
+        @return:
+        """
+        seq_length = len(solution) - 1
+        tsp_sequence = np.array(solution)
+        uncrosses = 0
+        for i in range(1, seq_length - 1):
+            for j in range(i + 1, seq_length):
+                new_tsp_sequence = TwoOpt.swap2opt(tsp_sequence, i, j)
+                new_distance = distance + TwoOpt.gain(i, j, tsp_sequence, matrix_dist)
+                if new_distance < distance:
+                    uncrosses += 1
+                    tsp_sequence = np.copy(new_tsp_sequence)
+                    distance = new_distance
+        return tsp_sequence, distance, uncrosses
+
+    @staticmethod
+    def swap2opt(tsp_sequence, i, j):
+        new_tsp_sequence = np.copy(tsp_sequence)
+        new_tsp_sequence[i:j + 1] = np.flip(tsp_sequence[i:j + 1], axis=0)  # flip or swap ?
+        return new_tsp_sequence
+
+    @staticmethod
+    def gain(i, j, tsp_sequence, matrix_dist):
+        old_link_len = (matrix_dist[tsp_sequence[i], tsp_sequence[i - 1]] + matrix_dist[
+            tsp_sequence[j], tsp_sequence[j + 1]])
+        changed_links_len = (matrix_dist[tsp_sequence[j], tsp_sequence[i - 1]] + matrix_dist[
+            tsp_sequence[i], tsp_sequence[j + 1]])
+        return - old_link_len + changed_links_len
+
+    @staticmethod
+    def loop2opt(solution, instance, max_num_of_uncrosses=400):  # Iterate step2opt max_iter times (2-opt local search)
+        """
+
+        @param tsp_sequence:
+        @param instance:
+        @param max_num_of_uncrosses:
+        @return:
+        """
+        matrix_dist = instance.dist_matrix
+        new_len = compute_lenght(solution, matrix_dist)
+        new_tsp_sequence = np.copy(np.array(solution))
+        uncross = 0
+        while uncross < max_num_of_uncrosses:
+            new_tsp_sequence, new_reward, uncr_ = TwoOpt.step2opt(new_tsp_sequence, matrix_dist, new_len)
+            uncross += uncr_
+            if new_reward < new_len:
+                new_len = new_reward
+            else:
+                return new_tsp_sequence.tolist(), new_len, uncross
+
+        return new_tsp_sequence.tolist(), new_len, uncross
+