KinDER: A Physical Reasoning Benchmark
for Robot Learning and Planning

Anonymous Submission. Under Review @ RSS 2026

Hover over the environments below to see them in action

BalanceBeam3D

BaseMotion3D

ClutteredRetrieval2D

ClutteredStorage2D

ConstrainedCupboard3D

DynObstruction2D

DynPushPullHook2D

DynPushT2D

DynScoopPour2D

Dynamo3D

Motion2D

Obstruction2D

Obstruction3D

Packing3D

PushPullHook2D

Rearrange3D

ScoopPour3D

Shelf3D

SortClutteredBlocks3D

StickButton2D

SweepIntoDrawer3D

SweepSimple3D

Table3D

Tossing3D

Transport3D

About KinDER

A physical reasoning benchmark for robot learning and planning.

Robotic systems that interact with the physical world must reason about kinematic and dynamic constraints imposed by their own embodiment, their environment, and the task at hand. We introduce KinDER, a benchmark for Kinematic and Dynamic Embodied Reasoning that targets physical reasoning challenges arising in robot learning and planning. KinDER comprises 25 procedurally generated environments, a Gymnasium-compatible Python library with parameterized skills and demonstrations, and a standardized evaluation suite with 8 implemented baselines spanning task and motion planning, imitation learning, reinforcement learning, and foundation-model-based approaches. The environments are designed to isolate five core physical reasoning challenges: basic spatial relations, nonprehensile multi-object manipulation, tool use, combinatorial geometric constraints, and dynamic constraints, disentangled from perception, language understanding, and application-specific complexity. Empirical evaluation shows that existing methods struggle to solve many of the environments, indicating substantial gaps in current approaches to physical reasoning. We additionally include real-to-sim-to-real experiments on a mobile manipulator to assess the correspondence between simulation and real-world physical interaction. KinDER is fully open-sourced and intended to enable systematic comparison across diverse paradigms for advancing physical reasoning in robotics.

What Makes KinDER Challenging? ▼

For Reinforcement Learning

Environments have long horizons and sparse rewards. Users are welcome to engineer dense rewards, but doing so may be nontrivial. Environments also have very diverse task distributions, so learned policies must generalize.

For Imitation Learning

Physical reasoning requires understanding physical constraints (e.g., spatial relationships, kinematics, dynamics). Imitating surface-level patterns in demonstrations is not enough to generalize to broad task distributions.

For Vision-Language Models

The physical reasoning required in KinDER is not easy to represent in natural language. Spatial reasonining, a subset of physical reasoning, is a known challenge for vision-language models.

For Hierarchical Approaches

Approaches that first decide "what to do" and then decide "how to do it" -- whether in hierarchical RL, planning, or with multi-level foundation models -- will run into difficulties when there are couplings between these high-level and low-level decisions.

For Task and Motion Planning

KinDER does not provide any models for TAMP. Users are welcome to engineer their own, but doing so may be nontrivial. Some environments contain many objects, which may make planning slow.

For Human Engineers

KinDER features diverse task distributions and long time horizons, making it challenging for engineers to hand-design even very environment-specific solutions. A single task may be straightforward, but designing generalized solutions is nontrivial.

KinDERGym: Usage

Installation

KinDER will be available on PyPI following publication. At that time, you will be able to install it with pip install kinder. Until then, you can install it from source as follows. First, we strongly recommend uv. Then choose one of the following based on your needs:

# Core dependencies only
# Everything (excluding develop)
uv pip install -e ".[all]"

# Specific environment categories
uv pip install -e ".[kinematic2d]"      # Kinematic 2D only
uv pip install -e ".[dynamic2d]"   # Dynamic 2D only
uv pip install -e ".[geom3d]"      # Kinematic 3D only
uv pip install -e ".[dynamic3d]"     # Dynamic 3D only

Basic Usage (Gym API)

import kinder
kinder.register_all_environments()
env = kinder.make("kinder/Obstruction2D-o3-v0")  # 3 obstructions
obs, info = env.reset()  # procedural generation
action = env.action_space.sample()
next_obs, reward, terminated, truncated, info = env.step(action)
img = env.render()

Object-Centric States ▼

All environments in KinDER use object-centric states:

from kinder.envs.geom2d.obstruction2d import ObjectCentricObstruction2DEnv
env = ObjectCentricObstruction2DEnv(num_obstructions=3)
obs, _ = env.reset(seed=123)
print(obs.pretty_str())

Here, obs is an ObjectCentricState, and the printout is:

############################################################### STATE ###############################################################
type: crv_robot           x         y    theta    base_radius    arm_joint    arm_length    vacuum    gripper_height    gripper_width
-----------------  --------  --------  -------  -------------  -----------  ------------  --------  ----------------  ---------------
robot              0.885039  0.803795  -1.5708            0.1          0.1           0.2         0              0.07             0.01

type: rectangle           x         y    theta    static    color_r    color_g    color_b    z_order      width     height
-----------------  --------  --------  -------  --------  ---------  ---------  ---------  ---------  ---------  ---------
obstruction0       0.422462  0.100001        0         0       0.75        0.1        0.1        100  0.132224   0.0766399
obstruction1       0.804663  0.100001        0         0       0.75        0.1        0.1        100  0.0805652  0.0955062
obstruction2       0.559246  0.100001        0         0       0.75        0.1        0.1        100  0.12608    0.180172

type: target_block          x         y    theta    static    color_r    color_g    color_b    z_order     width    height
--------------------  -------  --------  -------  --------  ---------  ---------  ---------  ---------  --------  --------
target_block          1.20082  0.100001        0         0   0.501961          0   0.501961        100  0.138302  0.155183

type: target_surface           x    y    theta    static    color_r    color_g    color_b    z_order     width    height
----------------------  --------  ---  -------  --------  ---------  ---------  ---------  ---------  --------  --------
target_surface          0.499675    0        0         1   0.501961          0   0.501961        101  0.180286       0.1
#####################################################################################################################################

For compatibility with baselines, observations are provided as vectors. Convert between vectors and object-centric states:

import kinder
kinder.register_all_environments()
env = kinder.make("kinder/Obstruction2D-o3-v0")
vec_obs, _ = env.reset(seed=123)
object_centric_obs = env.observation_space.devectorize(vec_obs)
recovered_vec_obs = env.observation_space.vectorize(object_centric_obs)