diff --git a/poetry.lock b/poetry.lock
index fd4321bde4..14fd4b94cd 100644
--- a/poetry.lock
+++ b/poetry.lock
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:04de7cc8e65d2c312801c6bea46a2e2778ffe21c1a60c1f8384d66306b03d485
-size 370444
+oid sha256:7616cdd0f1b405b1e867be8c62059002a1e6c7489ef213bf2b5673981614d58b
+size 369979
diff --git a/pyproject.toml b/pyproject.toml
index 4ceaef54ef..d9cc49bfd0 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -31,7 +31,7 @@ hexdump = "*"
 Jinja2 = "*"
 json-rpc = "*"
 libusb1 = "*"
-numpy = "~1.23"  # pinned for acados
+numpy = "*"
 onnx = ">=1.14.0"
 onnxruntime-gpu = { version = ">=1.15.1", platform = "linux", markers = "platform_machine == 'x86_64'" }
 pillow = "*"
diff --git a/selfdrive/controls/lib/lateral_planner.py b/selfdrive/controls/lib/lateral_planner.py
index baea407498..38258b4b5d 100644
--- a/selfdrive/controls/lib/lateral_planner.py
+++ b/selfdrive/controls/lib/lateral_planner.py
@@ -132,7 +132,7 @@ class LateralPlanner:
     lateralPlan.psis = self.lat_mpc.x_sol[0:CONTROL_N, 2].tolist()
 
     lateralPlan.curvatures = (self.lat_mpc.x_sol[0:CONTROL_N, 3]/self.v_ego).tolist()
-    lateralPlan.curvatureRates = [float(x/self.v_ego) for x in self.lat_mpc.u_sol[0:CONTROL_N - 1]] + [0.0]
+    lateralPlan.curvatureRates = [float(x.item() / self.v_ego) for x in self.lat_mpc.u_sol[0:CONTROL_N - 1]] + [0.0]
 
     lateralPlan.mpcSolutionValid = bool(plan_solution_valid)
     lateralPlan.solverExecutionTime = self.lat_mpc.solve_time
diff --git a/selfdrive/locationd/paramsd.py b/selfdrive/locationd/paramsd.py
index de1265f0d7..1644ceaf92 100755
--- a/selfdrive/locationd/paramsd.py
+++ b/selfdrive/locationd/paramsd.py
@@ -85,8 +85,8 @@ class ParamsLearner:
         # We observe the current stiffness and steer ratio (with a high observation noise) to bound
         # the respective estimate STD. Otherwise the STDs keep increasing, causing rapid changes in the
         # states in longer routes (especially straight stretches).
-        stiffness = float(self.kf.x[States.STIFFNESS])
-        steer_ratio = float(self.kf.x[States.STEER_RATIO])
+        stiffness = float(self.kf.x[States.STIFFNESS].item())
+        steer_ratio = float(self.kf.x[States.STEER_RATIO].item())
         self.kf.predict_and_observe(t, ObservationKind.STIFFNESS, np.array([[stiffness]]))
         self.kf.predict_and_observe(t, ObservationKind.STEER_RATIO, np.array([[steer_ratio]]))
 
@@ -198,14 +198,14 @@ def main(sm=None, pm=None):
         learner = ParamsLearner(CP, CP.steerRatio, 1.0, 0.0)
         x = learner.kf.x
 
-      angle_offset_average = clip(math.degrees(x[States.ANGLE_OFFSET]),
+      angle_offset_average = clip(math.degrees(x[States.ANGLE_OFFSET].item()),
                                   angle_offset_average - MAX_ANGLE_OFFSET_DELTA, angle_offset_average + MAX_ANGLE_OFFSET_DELTA)
-      angle_offset = clip(math.degrees(x[States.ANGLE_OFFSET] + x[States.ANGLE_OFFSET_FAST]),
+      angle_offset = clip(math.degrees(x[States.ANGLE_OFFSET].item() + x[States.ANGLE_OFFSET_FAST].item()),
                           angle_offset - MAX_ANGLE_OFFSET_DELTA, angle_offset + MAX_ANGLE_OFFSET_DELTA)
-      roll = clip(float(x[States.ROAD_ROLL]), roll - ROLL_MAX_DELTA, roll + ROLL_MAX_DELTA)
-      roll_std = float(P[States.ROAD_ROLL])
+      roll = clip(float(x[States.ROAD_ROLL].item()), roll - ROLL_MAX_DELTA, roll + ROLL_MAX_DELTA)
+      roll_std = float(P[States.ROAD_ROLL].item())
       # Account for the opposite signs of the yaw rates
-      sensors_valid = bool(abs(learner.speed * (x[States.YAW_RATE] + learner.yaw_rate)) < LATERAL_ACC_SENSOR_THRESHOLD)
+      sensors_valid = bool(abs(learner.speed * (x[States.YAW_RATE].item() + learner.yaw_rate)) < LATERAL_ACC_SENSOR_THRESHOLD)
       avg_offset_valid = check_valid_with_hysteresis(avg_offset_valid, angle_offset_average, OFFSET_MAX, OFFSET_LOWERED_MAX)
       total_offset_valid = check_valid_with_hysteresis(total_offset_valid, angle_offset, OFFSET_MAX, OFFSET_LOWERED_MAX)
       roll_valid = check_valid_with_hysteresis(roll_valid, roll, ROLL_MAX, ROLL_LOWERED_MAX)
@@ -215,8 +215,8 @@ def main(sm=None, pm=None):
       liveParameters = msg.liveParameters
       liveParameters.posenetValid = True
       liveParameters.sensorValid = sensors_valid
-      liveParameters.steerRatio = float(x[States.STEER_RATIO])
-      liveParameters.stiffnessFactor = float(x[States.STIFFNESS])
+      liveParameters.steerRatio = float(x[States.STEER_RATIO].item())
+      liveParameters.stiffnessFactor = float(x[States.STIFFNESS].item())
       liveParameters.roll = roll
       liveParameters.angleOffsetAverageDeg = angle_offset_average
       liveParameters.angleOffsetDeg = angle_offset
@@ -228,10 +228,10 @@ def main(sm=None, pm=None):
         0.2 <= liveParameters.stiffnessFactor <= 5.0,
         min_sr <= liveParameters.steerRatio <= max_sr,
       ))
-      liveParameters.steerRatioStd = float(P[States.STEER_RATIO])
-      liveParameters.stiffnessFactorStd = float(P[States.STIFFNESS])
-      liveParameters.angleOffsetAverageStd = float(P[States.ANGLE_OFFSET])
-      liveParameters.angleOffsetFastStd = float(P[States.ANGLE_OFFSET_FAST])
+      liveParameters.steerRatioStd = float(P[States.STEER_RATIO].item())
+      liveParameters.stiffnessFactorStd = float(P[States.STIFFNESS].item())
+      liveParameters.angleOffsetAverageStd = float(P[States.ANGLE_OFFSET].item())
+      liveParameters.angleOffsetFastStd = float(P[States.ANGLE_OFFSET_FAST].item())
       if DEBUG:
         liveParameters.filterState = log.LiveLocationKalman.Measurement.new_message()
         liveParameters.filterState.value = x.tolist()