← Paper Digest

When Will You Hand It Over? A Hidden Start Delay Makes Self-Play Handover Robust to Unfamiliar Partners

Mohammad Khoshnazar, Mike Degany, Deyuan Qu, Qing Yang, Matyas Szanto

November 2026

TL;DR A simulated Unitree G1 that learns to receive a parcel by PPO self-play succeeds in 90-95% of handovers with its own partner but only 31% with partners it has never met, because it learns when its partner starts. Adding a random hidden start delay to the partner raises success with unseen partners to 80% in both training seeds.

What they do

Cooperative self-play can teach a humanoid to take an object from a giver without any human data, but the trained pair may settle on a private convention. In a simulated parcel handover between a fixed-base Unitree G1 and a learned giver, the authors find that this convention is about when the giver starts. They fix it with one hidden rule on the giver.

How it works

  • Self-play. In Isaac Lab, the G1 receives a 0.5 kg parcel from a learned partner hand that sets its reach speed, small offer corrections and release. Both act from state, not images, and take turns being updated by PPO, with a pool of past partners.
  • Success measure. A supported transfer means the robot alone holds the parcel with a formed grip for 15 steps, with limits over the whole episode on unsupported steps, slip in the grasp and contact force between the robot and the partner’s hand.
  • Hidden start delay (HSD). Each episode draws a 0-1 s delay during which the partner’s hand must stay still. The receiver’s observations, rewards and network are unchanged, so it must read from the partner’s motion when the offer begins.

Results

  • Self-play receivers (two seeds) succeed 95.0 / 89.5% of the time with their own partner but 30.5 / 30.5% with scripted partners. Delaying their own partner by 1 s drops success from 95.0 to 6.0% and from 87.0 to 33.5%.
  • After 300 more updates with HSD, success with unseen scripted partners reaches 80.5 / 80.0%, against 35.0 / 41.0% for a control given the same extra training. Across eleven held-out partner conditions the mean rises from 30.5 / 27.2% to 78.8 / 62.7%.
  • Episodes with robot-partner contact above 20 N fall from 10.7 / 7.9% to 1.9 / 2.5%. Keeping the partner frozen gives the same robustness.

Limitations

Simulation only, with a simulated partner hand, state inputs instead of images and two training seeds. The scripted partners are designed behaviours, and their start times overlap the training delays. Every receiver stays below 5% with a scripted partner that withdraws if the parcel moves more than 3 cm before release.

Figures

From Figure 1: A Unitree G1 with a Dex3 hand reaches for the parcel, shares the load and holds it alone once the lear...
From Figure 1: A Unitree G1 with a Dex3 hand reaches for the parcel, shares the load and holds it alone once the learned partner lets go. The human character is an animation driven by the partner's hand.
From Figure 2: In training, receiver and partner act in one physics simulation and take turns being updated by PPO; h...
From Figure 2: In training, receiver and partner act in one physics simulation and take turns being updated by PPO; hidden partner rules sit between the partner's action and its hand controller. At inference only the receiver is kept and tested with partners it has never met.
From Figure 3: A scripted partner starts its offer about one second late. The self-play receiver (top) reaches at onc...
From Figure 3: A scripted partner starts its offer about one second late. The self-play receiver (top) reaches at once, presses on the rising parcel and never forms a grasp; after start-delay training (bottom) it takes the parcel as it arrives.
From Figure 4: (a) With its own learned partner, the self-play receiver collapses as the partner's hidden start delay...
From Figure 4: (a) With its own learned partner, the self-play receiver collapses as the partner's hidden start delay grows, while the start-delay-trained receiver stays flat. (b) Success with scripted partners on a fresh scenario seed; the late partners start later than any delay used in training.
Abstract

Cooperative self-play can teach a humanoid to receive an object without any human data, but the partner it meets after training is never the one it trained with; we show that a fluent self-play handover policy breaks with unfamiliar partners, identify the main cause, and repair it. In a physics-based parcel handover between a Unitree G1 and a learned giver, both trained by alternating PPO self-play, the receiver completes 95% and 90% of handovers with its own partner (two training seeds) but only 31% with scripted partners it has never met. Paired interventions isolate the cause: under a shared reward the learned giver always starts its offer at once, so the receiver learns a single rendezvous, and delaying its own partner by one second cuts success from 95% to 6% (87% to 34% in the second seed). We introduce hidden start-delay randomisation (HSD), a one-line rule on the learned partner that the receiver never observes. After 300 further updates, success with unseen scripted partners rises from 31% to 80% in both seeds, against 35--41% for a budget-matched control. The gain holds on a fresh scenario seed, on eleven held-out partner conditions and on partners that start later than any training delay; it cuts high-force contact three- to five-fold and needs no retraining of the partner.

Summaries are written automatically from each paper's text. Figures and videos come from the paper and its project page; all credit goes to the authors.