CARE: Consistent Action-Response Execution for Human-Guided Robot Transport
Authors: Ning Liu, Sen Shen, Zheng Li, Matthew D'Souza, Jen Jen Chung, Lan Wu, Thomas Bräunl
Organizations: School of Engineering, The University of Western Australia, 35 Stirling Highway, Perth, 6009, WA, Australia · Department of Mechanical and Automation Engineering, The Chinese University of Hong Kong, Shatin, New Territories, Hong Kong SAR · School of Electrical Engineering and Computer Science, The University of Queensland, Brisbane, 4072, QLD, Australia
Human-guided collaborative transport becomes interaction-critical when the operator can continue along a route segment that is not executable by the robot--payload system. In this regime, operator motion is not an exogenous reference: it simultaneously communicates a persistent route tendency and adapts to the robot's realized progress. A common modular formulation predicts one operator future from observation history, scores multiple robot proposals against that shared forecast, and then modifies the selected command for local feasibility. This can compare candidates under responses they would not induce and can condition prediction on an action that is never executed. We formulate local control over executable-action-indexed joint human--robot transitions and instantiate it in Consistent Action--Response Execution (CARE). Each nominal proposal or bounded alternative is first mapped to an executable action; that same action conditions the route-aligned operator-response prediction, the matched robot--operator rollout, and a fixed interaction utility for candidate selection. Bounded reciprocal refinement handles transient route--feasibility mismatch, whereas persistent progress loss replaces the nominal proposal with a progress-oriented recovery proposal while retaining the same downstream executable-action-conditioned evaluation. Across training and held-out maps, CARE preserves near-saturated performance when the indicated route remains locally usable and achieves the highest observed success and service availability when the operator can continue but the robot must detour. It also yields lower jerk-proxy values and measured controller latency than dynamic-window and Timed Elastic Band baselines. Factorial ablations distinguish reciprocal refinement from progress recovery, delayed-response tests probe temporal mismatch, and an indoor deployment demonstrates physical executability.
Figures & tables
Figure 1: Physical human-guided transport with CARE. The red curve shows the executed robot trajectory while negotiating a local obstacle
Figure 2: Schematic overview of the CARE closed loop. The progress supervisor selects the nominal or recovery proposal, after which candidate actions undergo realization, operator-response prediction, matched rollout, and interaction evaluation before execution. Simplified action notation is used for readability; the formal action realization is defined in Sec. 3.1.1 , while the candidate construction and improvement-margin selection rules are given in Eqs. ( 8 )–( 12 ).
Side
Route
Method
Succ. ↑
Serv. ↑
Jerk prox. ↓
Avg. Lat. ↓
Succ. ↑
Serv. ↑
Jerk prox. ↓
Avg. Lat. ↓
(%)
(%)
(a.u./s 2 )
(ms)
(%)
(%)
(a.u./s 2 )
(ms)
Training maps
MPC
95.33
96.76 14.74
2.92 6.03
6.62 2.93
95.67
95.89 14.65
3.16 3.36
6.63 2.92
ORCA
96.67
97.68 12.56
8.00 1.98
5.53 1.75
75.00
75.89 32.58
6.86 3.19
5.69 1.76
TEB
95.67
97.01 13.95
14.98 4.99
13.99 5.90
93.00
92.47 17.55
17.81 5.22
14.69 6.22
Table 1: Closed-loop performance on training and held-out Side and Route maps over 300 seed-matched episodes per method–condition cell. Continuous metrics report mean ± standard deviation. Best results within each map split and condition are shown in bold.
Modules
Side
Route
Recip.
Recov.
Succ. ↑
Comfort viol. ↓
Wait ↓
Succ. ↑
Comfort viol. ↓
Wait ↓
(%)
(%)
(s)
(%)
(%)
(s)
✗
✗
100.00
6.89 ± 6.96
0.32 ± 1.09
96.33
5.99 ± 6.58
1.71 ± 11.46
✓
✗
100.00
5.18 ± 5.36
0.24 ± 0.77
97.67
5.02 ± 5.46
1.63 ± 11.57
✗
✓
100.00
6.91 ± 6.94
0.27 ± 1.04
98.33
6.11 ± 6.60
0.41 ± 1.24
✓
✓
100.00
4.79 ± 5.34
0.21 ± 0.61
98.67
4.53 ± 5.46
0.38 ± 1.69
Table 2: Factorial ablation of reciprocal refinement and progress recovery on held-out Side and Route maps, with 300 seed-matched episodes per configuration–condition cell. Best results within each condition are shown in bold.
Figure 3: Effect of response conditioning on reciprocal activation, realization exposure, and selected realization gap over held-out Route episodes. Error bars show bootstrap 95% confidence intervals
Response conditioning
Succ. ↑
Serv. ↑
Wait ↓
(%)
(%)
(s)
Shared
97.67
97.72 ± 4.84
0.81 ± 2.31
Raw-conditioned
98.33
98.87 ± 3.89
0.67 ± 1.98
Executable-action (CARE)
98.67
99.36 ± 3.01
0.38 ± 1.69
Table 3: Task-level effect of response conditioning on held-out Route maps over 300 seed-matched episodes. Continuous metrics report mean ± standard deviation. Best results are shown in bold.
Figure 4: Cumulative observation-representation study for the nominal policy over pooled held-out Side and Route episodes. The four layouts are independently trained using the same training pipeline and evaluated under the same matched-seed protocol. Error bars show bootstrap 95% confidence intervals
Figure 5: Robustness to operator-response delay on held-out Route maps. Error bars show two-sided Wilson 95% confidence intervals
Figure 6: Robustness to LiDAR sensing range on held-out Route maps. Panels report task success and operator wait. Error bars show two-sided Wilson 95% confidence intervals for success and episode-level percentile-bootstrap 95% confidence intervals for operator wait
We focus on human-robot collaborative transport, a challenging task of broad relevance spanning logistics, manufacturing, and the home, in which a user and a robot work together to relocate a large or heavy object. To act as an effective partner, the robot should reduce the user's effort by contributing to efficient relocation of the object while remaining physically responsive to them. Prior work often addresses these capabilities separately, producing robots that may move the object efficiently but resist user input, or accommodate the user but depend on continuous guidance. Our key insight is that obstacle-constrained collaborative transport requires integrating predictions of human collaborative behavior with compliant robot control. To this end, we introduce PROACT, a framework for human-robot collaborative transport that incorporates anticipation into compliant whole-body control through a learned model of human collaborative behavior. Trained on a large-scale, real-world dataset of dyadic human transport demonstrations, our transformer architecture distills collaborative behavior into predictions of future object motion. Across 108 real-world trials with a 9-DoF mobile manipulator, PROACT reduces mean interaction work by 59.2% and 20.4%, and mean completion time by 12.9% and 6.9%, relative to compliance-only and MPC baselines, respectively. Footage from our experiments can be found at https://youtu.be/qAGvQfVPjbk.
Elvin Yang, Christoforos Mavrogiannis
Department of Robotics, University of Michigan, Ann Arbor, United States
Demonstration-conditioned policies provide a natural interface for specifying robot behavior, yet long-horizon manipulation remains difficult when visually similar states recur across different stages or when demonstrations and executions proceed at different speeds. We identify the resulting failure mode as stage confusion and introduce Progress-Aligned Context for Execution (PACE), a stateful method that continually reinterprets a complete demonstration according to realized execution progress. PACE compresses the demonstration into ordered multimodal prompt tokens and uses training-only dual-edge attention supervision to expose its latent stage structure. During execution, an episode-local fast-weight memory causally encodes realized action-observation transitions and modulates prompt cross-attention, producing a progress-aligned context for a unified diffusion action expert without test-time stage labels or stage-specific policies. PACE improves success from 88.9% to 94.0% on LIBERO-Gen Goal Chain, from 79.1% to 83.3% on Spatial Combination, and from 33.3% to 73.3% on the two-step Block Routing tasks. Failure analysis further indicates that structured demonstration alignment and causal execution memory jointly mitigate stage confusion.
Yenan Chen, Junjie Shi, Lu Chen +2
College of Control Science and Engineering, Zhejiang University, Hangzhou 310027, China
Recent vision-language-action and diffusion-based robot policies often use action chunking, where each policy query predicts a sequence of future actions and the robot executes an open-loop prefix before re-querying. While this interface improves local motion continuity, deployment still requires choosing the execution horizon: how much of each predicted chunk should be executed before acquiring a new observation. However, our experiments show that success is strongly task-dependent and non-monotonic with respect to the execution horizon, making a single constant horizon an unreliable deployment rule. We propose PACE (Phase-Aware Chunk Execution), a training-free test-time execution method that selects the execution horizon online from the predicted chunk itself. PACE exploits the phase-dependent kinematic structure of manipulation trajectories by identifying low-speed transition points in the predicted speed profile and using them as candidate replanning boundaries. Because PACE uses only the predicted action chunk, it is plug-and-play and requires no retraining or access to policy internals. We validate PACE through large-scale evaluations in both simulation and real-robot settings. On 50 RoboTwin2.0 tasks, PACE raises the average success rate from 57.8% to 64.2%. In real-robot experiments on bimanual ALOHA and single-arm Franka platforms, PACE improves the average task score from 60.7 to 77.7 and the average success rate from 50.7% to 70.4%. Ablations and rollout-level analyses show that PACE adapts execution horizons across manipulation phases, shortening near transitions while preserving longer execution during coherent motion.
Junnan Nie, Jiayi Li, Chenghao Liu +5
Peking University · Peking University. · JD Explore Academy. +1