diff --git a/frc971/control_loops/python/BUILD b/frc971/control_loops/python/BUILD
index d172983..897738a 100644
--- a/frc971/control_loops/python/BUILD
+++ b/frc971/control_loops/python/BUILD
@@ -26,3 +26,13 @@
     ':controls',
   ],
 )
+
+py_binary(
+  name = 'down_estimator',
+  srcs = [
+    'down_estimator.py',
+  ],
+  deps = [
+    ':controls',
+  ],
+)
diff --git a/frc971/control_loops/python/down_estimator.py b/frc971/control_loops/python/down_estimator.py
new file mode 100644
index 0000000..1d776b3
--- /dev/null
+++ b/frc971/control_loops/python/down_estimator.py
@@ -0,0 +1,104 @@
+#!/usr/bin/python
+
+import math
+import sys
+import random
+
+import numpy
+import gflags
+import glog
+from matplotlib import pylab
+
+from frc971.control_loops.python import controls
+
+FLAGS = gflags.FLAGS
+
+gflags.DEFINE_bool('plot', False, 'If true, plot the loop response.')
+
+class DownEstimator(control_loop.ControlLoop):
+  def __init__(self, name='DownEstimator'):
+    super(DownEstimator, self).__init__(name)
+    self.dt = 0.005
+
+    # State is [gyro_angle, bias].
+    # U is [gyro_x_velocity].
+
+    self.A_continuous = numpy.matrix([[0, 1],
+                                      [0, 0]])
+
+    self.B_continuous = numpy.matrix([[1],
+                                      [0]])
+
+    self.A, self.B = c2d(self.A_continuous, self.B_continuous, dt):
+
+    q_gyro_noise = 1e-6
+    self.Q = numpy.matrix([[1.0 / (q_gyro_noise ** 2.0), 0.0],
+                           [0.0, 1.0 / (q_gyro_noise ** 2.0)]])
+
+    r_accelerometer_angle_noise = 5e-3
+    self.R = numpy.matrix([[(r_accelerometer_angle_noise ** 2.0)]])
+
+    self.C = numpy.matrix([[1.0, 0.0]])
+
+    self.KalmanGain, self.Q_steady = controls.kalman(
+        A=self.A, B=self.B, C=self.C, Q=self.Q, R=self.R)
+
+    self.L = self.A * self.KalmanGain
+
+    self.InitializeState()
+
+  def Update(self, accelerometer_x, accelerometer_y, accelerometer_z, gyro_x):
+    acceleration = math.sqrt(
+        accelerometer_x**2 + accelerometer_y**2 + accelerometer_z**2)
+    if acceleration < 0.9 or acceleration > 1.1:
+      glog.error('bad total acceleration %f' % acceleration)
+      # TODO(Brian): Tune this?
+      return
+    accelerometer_angle = math.atan2(accelerometer_x, accelerometer_z)
+    Z = numpy.matrix([[accelerometer_angle], [gyro_x]])
+
+    Y = Z - self.H * self.X_hat
+    S = self.H * self.P * self.H.transpose() + self.R
+    K = self.P * self.H.transpose() * numpy.linalg.inv(S)
+    self.X_hat += K * Y
+    self.P = (numpy.identity(K.shape[0]) - K * self.H) * self.P
+
+def main(argv):
+  argv = FLAGS(argv)
+  glog.init()
+
+  estimator = DownEstimator()
+
+  if FLAGS.plot:
+    real_angles = [0]
+    real_velocities = [0]
+    estimated_angles = [0]
+    estimated_velocities = [0]
+    for _ in xrange(100):
+      estimator.Predict(0)
+      estimator.Update(numpy.sqrt(2) / 2.0, numpy.sqrt(2) / 2.0, 0, 0)
+      real_angles.append(math.pi / 2)
+      real_velocities.append(0)
+      estimated_angles.append(estimator.X_hat[0, 0])
+      estimated_velocities.append(estimator.X_hat[1, 0])
+    angle = math.pi / 2
+    velocity = 1
+    for i in xrange(100):
+      measured_velocity = velocity + (random.random() - 0.5) * 0.01 + 0.05
+      estimator.Predict(measured_velocity)
+      estimator.Update(math.sin(angle) + (random.random() - 0.5) * 0.02, 0,
+                       math.cos(angle) + (random.random() - 0.5) * 0.02,
+                       measured_velocity)
+      real_angles.append(angle)
+      real_velocities.append(velocity)
+      estimated_angles.append(estimator.X_hat[0, 0])
+      estimated_velocities.append(estimator.X_hat[1, 0])
+      angle += velocity * estimator.dt
+    pylab.plot(range(len(real_angles)), real_angles)
+    pylab.plot(range(len(real_velocities)), real_velocities)
+    pylab.plot(range(len(estimated_angles)), estimated_angles)
+    pylab.plot(range(len(estimated_velocities)), estimated_velocities)
+    pylab.show()
+
+if __name__ == '__main__':
+  sys.exit(main(sys.argv))