diff --git a/package.xml b/package.xml
index 7986b46..c825d4d 100644
--- a/package.xml
+++ b/package.xml
@@ -16,4 +16,5 @@
     <depend>geometry_msgs</depend>
     <depend>rospy</depend>
     <depend>std_msgs</depend>
+    <depend>trac_ik</depend>
 </package>
diff --git a/src/active_grasp/baselines.py b/src/active_grasp/baselines.py
index 3b4049e..4f05ff7 100644
--- a/src/active_grasp/baselines.py
+++ b/src/active_grasp/baselines.py
@@ -13,14 +13,14 @@ class TopView(SingleViewPolicy):
     def activate(self, bbox, view_sphere):
         super().activate(bbox, view_sphere)
         self.x_d = self.view_sphere.get_view(0.0, 0.0)
-        self.done = False if self.is_feasible(self.x_d) else True
+        self.done = False if self.solve_cam_ik(self.q0, self.x_d) else True
 
 
 class TopTrajectory(MultiViewPolicy):
     def activate(self, bbox, view_sphere):
         super().activate(bbox, view_sphere)
         self.x_d = self.view_sphere.get_view(0.0, 0.0)
-        self.done = False if self.is_feasible(self.x_d) else True
+        self.done = False if self.solve_cam_ik(self.q0, self.x_d) else True
 
     def update(self, img, x, q):
         self.integrate(img, x, q)
diff --git a/src/active_grasp/nbv.py b/src/active_grasp/nbv.py
index 24964c6..4f986b1 100644
--- a/src/active_grasp/nbv.py
+++ b/src/active_grasp/nbv.py
@@ -104,7 +104,7 @@ class NextBestView(MultiViewPolicy):
         view_candidates = []
         for theta, phi in itertools.product(thetas, phis):
             view = self.view_sphere.get_view(theta, phi)
-            if self.is_feasible(view, q):
+            if self.solve_cam_ik(q, view):
                 view_candidates.append(view)
         return view_candidates
 
diff --git a/src/active_grasp/policy.py b/src/active_grasp/policy.py
index 0d06996..90b320c 100644
--- a/src/active_grasp/policy.py
+++ b/src/active_grasp/policy.py
@@ -2,20 +2,27 @@ import numpy as np
 from sensor_msgs.msg import CameraInfo
 from pathlib import Path
 import rospy
+from trac_ik_python.trac_ik import IK
+
 
 from .timer import Timer
 from .rviz import Visualizer
-from robot_helpers.model import KDLModel
 from robot_helpers.ros import tf
 from robot_helpers.ros.conversions import *
 from vgn.detection import *
 from vgn.perception import UniformTSDFVolume
 
 
+def solve_ik(q0, pose, solver):
+    x, y, z = pose.translation
+    qx, qy, qz, qw = pose.rotation.as_quat()
+    return solver.get_ik(q0, x, y, z, qx, qy, qz, qw)
+
+
 class Policy:
     def __init__(self):
         self.load_parameters()
-        self.init_robot_model()
+        self.init_ik_solver()
         self.init_visualizer()
 
     def load_parameters(self):
@@ -28,18 +35,20 @@ class Policy:
         self.intrinsic = from_camera_info_msg(msg)
         self.qual_thresh = rospy.get_param("vgn/qual_threshold")
 
-    def init_robot_model(self):
-        self.model = KDLModel.from_parameter_server(self.base_frame, self.cam_frame)
-        self.ee_model = KDLModel.from_parameter_server(self.base_frame, "panda_link8")
+    def init_ik_solver(self):
+        self.q0 = [0.0, -0.79, 0.0, -2.356, 0.0, 1.57, 0.79]
+        self.cam_ik_solver = IK(self.base_frame, self.cam_frame)
+        self.ee_ik_solver = IK(self.base_frame, "panda_link8")
+
+    def solve_cam_ik(self, q0, view):
+        return solve_ik(q0, view, self.cam_ik_solver)
+
+    def solve_ee_ik(self, q0, pose):
+        return solve_ik(q0, pose, self.ee_ik_solver)
 
     def init_visualizer(self):
         self.vis = Visualizer()
 
-    def is_feasible(self, view, q_init=None):
-        if q_init is None:
-            q_init = [0.0, -0.79, 0.0, -2.356, 0.0, 1.57, 0.79]
-        return self.model.ik(q_init, view) is not None
-
     def activate(self, bbox, view_sphere):
         self.vis.clear()
 
@@ -82,7 +91,7 @@ class Policy:
             tip = pose.rotation.apply([0, 0, 0.05]) + pose.translation
             if self.bbox.is_inside(tip):
                 grasp.pose = pose
-                q_grasp = self.ee_model.ik(q, pose * self.T_grasp_ee)
+                q_grasp = self.solve_ee_ik(q, pose * self.T_grasp_ee)
                 if q_grasp is not None:
                     filtered_grasps.append(grasp)
                     scores.append(self.score_fn(grasp, quality, q, q_grasp))