diff --git a/libs/utils/trace.py b/libs/utils/trace.py
index f1cff7e3d07faf75f9b9ab26298fc7469dd1f135..39f6269374e15f6394aefd7fec283d2069c187fe 100644
--- a/libs/utils/trace.py
+++ b/libs/utils/trace.py
@@ -51,6 +51,10 @@ class Trace(object):
         # Maximum timespan for all collected events
         self.time_range = 0
 
+        # Time the system was overutilzied
+        self.overutilized_time = 0
+        self.overutilized_prc = 0
+
         # The dictionary of tasks descriptors available in the dataset
         self.tasks = {}
 
@@ -147,6 +151,15 @@ class Trace(object):
         logging.info('Collected events spans a %.3f [s] time interval',
                 self.time_range)
 
+        # Build a stat on trace overutilization
+        if self.hasEvents('sched_overutilized'):
+            df = self.df('sched_overutilized')
+            self.overutilized_time = df[df.overutilized == 1].len.sum()
+            self.overutilized_prc = 100. * self.overutilized_time / self.time_range
+
+            logging.info('Overutilized time: %.6f [s] (%.3f%% of trace time)',
+                    self.overutilized_time, self.overutilized_prc)
+
     def getTasks(self, dataframe=None,
             task_names=None, name_key='comm', pid_key='pid'):
         # """ Helper function to get PIDs of specified tasks
@@ -305,3 +318,11 @@ class Trace(object):
             [ccol > 2e9, ccol > 0, ccol > -2e9],
             ['Optimal Accept', 'SchedTune Accept', 'SchedTune Reject'], 'Suboptimal Reject')
 
+    def _sanitize_SchedOverutilized(self):
+        if not self.hasEvents('sched_overutilized'):
+            return
+        # Add a column with overutilized status duration
+        df = self.df('sched_overutilized')
+        df['start'] = df.index
+        df['len'] = (df.start - df.start.shift()).fillna(0).shift(-1)
+        df.drop('start', axis=1, inplace=True)
diff --git a/libs/utils/trace_analysis.py b/libs/utils/trace_analysis.py
index 35dfa4a4cca0840c1662b1883f3ac4d171668fd7..866705d503f8046b63c9ef1843e170d7f09f01e2 100644
--- a/libs/utils/trace_analysis.py
+++ b/libs/utils/trace_analysis.py
@@ -128,6 +128,7 @@ class TraceAnalysis(object):
         axes.grid(True);
         axes.set_xticklabels([])
         axes.set_xlabel('')
+        self.plotOverutilized(axes)
 
         axes = pltaxes[1]
         axes.set_title('LITTLE Cluster');
@@ -143,6 +144,7 @@ class TraceAnalysis(object):
         axes.set_xlim(self.x_min, self.x_max);
         axes.set_ylabel('MHz')
         axes.grid(True);
+        self.plotOverutilized(axes)
 
         # Save generated plots into datadir
         figname = '{}/cluster_freqs.png'.format(self.plotsdir)
@@ -253,6 +255,32 @@ class TraceAnalysis(object):
         lcpus = set(cpus) & set(self.platform['clusters']['little'])
         self.__plotCPU(lcpus, "LITTLE")
 
+    def plotOverutilized(self, axes=None):
+        if not self.trace.hasEvents('sched_overutilized'):
+            logging.warn('Events [sched_overutilized] not found, '\
+                    'plot DISABLED!')
+            return
+
+        # Build sequence of overutilization "bands"
+        df = self.trace.df('sched_overutilized')
+        bands = [(t, df['len'][t], df['overutilized'][t]) for t in df.index]
+
+        # If not axis provided: generate a standalone plot
+        if not axes:
+            gs = gridspec.GridSpec(1, 1)
+            plt.figure(figsize=(16, 1))
+            axes = plt.subplot(gs[0,0])
+            axes.set_title('System Status {white: EAS mode, red: Non EAS mode}');
+            axes.set_xlim(self.x_min, self.x_max);
+            axes.grid(True);
+
+        # Otherwise: draw overutilized bands on top of the specified plot
+        for (t1,td,overutilized) in bands:
+            if not overutilized:
+                continue
+            t2 = t1+td
+            axes.axvspan(t1, t2, facecolor='r', alpha=0.1)
+
     def plotTasks(self, tasks=None):
         if not self.trace.hasEvents('sched_load_avg_task'):
             logging.warn('Events [sched_load_avg_task] not found, '\
@@ -307,6 +335,7 @@ class TraceAnalysis(object):
             axes.grid(True);
             axes.set_xticklabels([])
             axes.set_xlabel('')
+            self.plotOverutilized(axes)
 
             # Plot CPUs residency
             axes = plt.subplot(gs[1,0]);
@@ -323,6 +352,7 @@ class TraceAnalysis(object):
             axes.grid(True);
             axes.set_xticklabels([])
             axes.set_xlabel('')
+            self.plotOverutilized(axes)
 
             # Plot PELT signals
             axes = plt.subplot(gs[2,0]);
@@ -331,6 +361,7 @@ class TraceAnalysis(object):
             data.plot(ax=axes, drawstyle='steps-post');
             axes.set_xlim(self.x_min, self.x_max);
             axes.grid(True);
+            self.plotOverutilized(axes)
 
             # Save generated plots into datadir
             figname = '{}/task_util_{}.png'.format(self.plotsdir, task_name)
@@ -401,6 +432,7 @@ class TraceAnalysis(object):
         axes.grid(True);
         axes.set_xticklabels([])
         axes.set_xlabel('')
+        self.plotOverutilized(axes)
 
         # Plot2: energy and capacity variations
         axes = plt.subplot(gs[1,:]);
@@ -416,6 +448,7 @@ class TraceAnalysis(object):
         axes.grid(True);
         axes.set_xticklabels([])
         axes.set_xlabel('')
+        self.plotOverutilized(axes)
 
         # Plot3: energy payoff
         axes = plt.subplot(gs[2,:]);
@@ -430,12 +463,15 @@ class TraceAnalysis(object):
         axes.grid(True);
         axes.set_xticklabels([])
         axes.set_xlabel('')
+        self.plotOverutilized(axes)
 
         # Plot4: energy deltas (kernel and host computed values)
         axes = plt.subplot(gs[3,:]);
         axes.set_title('Energy Deltas Values');
         df[['nrg_delta', 'nrg_diff_pct']].plot(ax=axes, style=['ro', 'b+']);
         axes.set_xlim(self.x_min, self.x_max);
+        axes.grid(True);
+        self.plotOverutilized(axes)
 
         # Save generated plots into datadir
         figname = '{}/ediff_time.png'.format(self.plotsdir)