diff --git a/include/cv/aggregator.hpp b/include/cv/aggregator.hpp
index 003173a9..c10f8ee2 100644
--- a/include/cv/aggregator.hpp
+++ b/include/cv/aggregator.hpp
@@ -53,7 +53,9 @@ class CVAggregator {
 
     // gets the record of all cv results
     LockPtr<std::map<int, IdentifiedTarget>> getCVRecord();
-
+    // Calculates the clustered center of each airdroptype in matched results,
+    // then assigns matched_results.matched_airdrop to the results
+    void calculateClusters();
     void updateRecords(std::vector<IdentifiedTarget>& new_values);
 
  private:
diff --git a/src/cv/aggregator.cpp b/src/cv/aggregator.cpp
index 1e61b098..346a0cb1 100644
--- a/src/cv/aggregator.cpp
+++ b/src/cv/aggregator.cpp
@@ -4,6 +4,7 @@
 #include "utilities/lockptr.hpp"
 #include "utilities/locks.hpp"
 #include "utilities/logging.hpp"
+#include "cv/clustering.hpp"
 
 CVAggregator::CVAggregator(Pipeline&& p) : pipeline(std::move(p)) {
     this->num_worker_threads = 0;
@@ -123,3 +124,35 @@ std::vector<AggregatedRun> CVAggregator::popAllRuns() {
     this->results->runs.clear();
     return out;
 }
+
+void CVAggregator::calculateClusters() {
+    LockPtr<std::map<int, IdentifiedTarget>> results = this->getCVRecord();
+        std::vector<std::vector<GPSCoord>> cluster_input;
+        for (const auto& pair : *results.data.get()) {
+            GPSProtoVec cords = pair.second.coordinates();
+            auto types = pair.second.target_type();
+            // add all types up to the current, in order to keep list format
+            for (int i = 0; i < cords.size(); i++) {
+                GPSCoord location = cords.at(i);
+                int type = types.at(i);
+                if (cluster_input.size() < type) {
+                    for (int j = cluster_input.size(); j < type; j++) {
+                        cluster_input.push_back(std::vector<GPSCoord>());
+                    }
+                }
+                cluster_input[type].push_back(location);
+            }
+        }
+        Clustering clustering;
+        std::vector<GPSCoord> clusterCenters = clustering.FindClustersCenter(cluster_input);
+
+        LockPtr<MatchedResults> matched = this->getMatchedResults();
+        std::unordered_map<AirdropType, AirdropTarget> matched_clusters;
+        for (int i = 0; i < clusterCenters.size(); i++) {
+             AirdropTarget airdrop;
+             airdrop.set_index(static_cast<AirdropType>(i));
+             airdrop.mutable_coordinate()->CopyFrom(clusterCenters[i]);
+             matched_clusters.insert(std::pair(static_cast<AirdropType>(i), std::move(airdrop)));
+        }
+        matched.data->matched_airdrop = std::move(matched_clusters);
+}
diff --git a/src/ticks/cv_loiter.cpp b/src/ticks/cv_loiter.cpp
index ccd6d4e4..f16ebe35 100644
--- a/src/ticks/cv_loiter.cpp
+++ b/src/ticks/cv_loiter.cpp
@@ -6,6 +6,7 @@
 #include "ticks/fly_search.hpp"
 #include "ticks/ids.hpp"
 #include "utilities/constants.hpp"
+#include "cv/clustering.hpp"
 
 CVLoiterTick::CVLoiterTick(std::shared_ptr<MissionState> state) : Tick(state, TickID::CVLoiter) {
     this->status = CVLoiterTick::Status::None;
@@ -39,9 +40,7 @@ Tick* CVLoiterTick::tick() {
             AirdropType::Water, AirdropType::Beacon,
         };
         */
-
-        LockPtr<CVResults> results = state->getCV()->getResults();
-
+       state->getCV()->calculateClusters();
         // for (const auto& bottle : ALL_BOTTLES) {
         //     // contains will never be false but whatever
         //     if (results.data->matches.contains(bottle)) {
diff --git a/tests/integration/clustering.cpp b/tests/integration/clustering.cpp
index 0bb638af..45ef9344 100644
--- a/tests/integration/clustering.cpp
+++ b/tests/integration/clustering.cpp
@@ -6,7 +6,7 @@
 #define POINT_DISTANCE 10
 #define NUM_OUTLIERS 5
 #define NUM_RUNS 5
-//redeclaring method to not have fixed seed so the test is deterministic 
+//redeclaring method to not have fixed seed so the test is non-deterministic 
 unsigned int seed = time(NULL);
 double random(double min, double max) {
     return min + static_cast<double>(rand_r(&seed)) / RAND_MAX * (max - min);
@@ -86,11 +86,14 @@ int main()
         scatter(x, y, std::vector<double>(), c);
         hold(on);
         auto plot = scatter(cluster_x, cluster_y);
+        ::matplot::legend({"Cluster points", "Calculated cluster Center"});
         plot->marker_style(line_spec::marker_style::diamond);
+        title("Clustering run " + std::to_string(run));
         std::ostringstream stringStream;
         stringStream << "run" << run << ".png";
         std::string copyOfStr = stringStream.str();
-     //  std::filesystem::remove(copyOfStr);
+        std::cout << "saved to /build/" << copyOfStr << std::endl;
+        //std::filesystem::remove(copyOfStr);
         save(copyOfStr);
         hold(off);