Deliberate Practice: Learning Robot Skills under a Budget

Shivam Vats, Sudarshan Harithas, Mete Tuluhan Akbulut, Arvind Raghunathan, George Konidaris · N/A · 2026

Summary

We consider the problem of autonomously learning robot skills under a limited practice budget for sequential tasks.

Abstract Summary

We consider the problem of autonomously learning robot skills under a limited practice budget for sequential tasks. We propose an active skill learning algorithm, \emph{Deliberate Practice (DP)}, that computes a provably \emph{budget-optimal} allocation---practicing skills that maximize expected cumulative reward while being learnable within the budget. DP estimates both the time needed to master skills and the cumulative reward of the task plans that the skills unlock. Computing a budget-optimal allocation is challenging as it requires reasoning about combinatorially many skill plans over a large practice budget. Our key contribution is a bilinear program that can compute this exactly using off-the-shelf solvers. Through simulated and real-world experiments on long-horizon manipulation tasks, we show that our approach allows robots to optimally use limited practice time to acquire useful policies and improve long-horizon planning.

Key Points

We consider the problem of autonomously learning robot skills under a limited practice budget for sequential tasks.
We propose an active skill learning algorithm, \emph{Deliberate Practice (DP)}, that computes a provably \emph{budget-optimal} allocation---practicing skills that maximize expected cumulative reward while being learnable within the budget.
DP estimates both the time needed to master skills and the cumulative reward of the task plans that the skills unlock.
Computing a budget-optimal allocation is challenging as it requires reasoning about combinatorially many skill plans over a large practice budget.
Our key contribution is a bilinear program that can compute this exactly using off-the-shelf solvers

title: ‘Deliberate Practice: Learning Robot Skills under a Budget’ slug: deliberate-practice-learning-robot-skills-under-a-budget authors:

Shivam Vats
Sudarshan Harithas
Mete Tuluhan Akbulut
Arvind Raghunathan
George Konidaris venue: arXiv preprint year: 2026 institution: N/A month: 8 bibtex: “@article{Vats2026_260813415v1,\n title = {Deliberate Practice: Learning
\ Robot Skills under a Budget},\n author = {Shivam Vats and Sudarshan Harithas
\ and Mete Tuluhan Akbulut and Arvind Raghunathan and George Konidaris},\n year
\ = {2026},\n eprint = {2608.13415v1},\n archivePrefix = {arXiv},\n primaryClass
\ = {cs.RO},\n url = {https://arxiv.org/abs/2608.13415v1}\n}” topics:
manipulation
reinforcement-learning
planning summary: We consider the problem of autonomously learning robot skills under a limited practice budget for sequential tasks. We propose an active skill learning algorithm, \emph{Deliberate Practice (DP)}, that computes a provably \emph{budget-optimal} allocation---

Budgeted Skill Learning for Planning

Given the high-level descriptions of a library of skills $\mathcal{U}$, consisting of their preconditions, termination conditions, and effects, the robot must learn the policy parameters of its skills to improve overall planning performance under a limited practice budget $B$, e.g., maximum number of trials. We adopt the standard episodic reinforcement learning setting without access to expert demonstrations.

Our goal is to compute a budget allocation $b_{u}$ across all skills $\mathcal{U}$ to maximize the expected planning performance.

Deliberate Practice

Our algorithm allocates the practice budget provably optimally by jointly reasoning about all robot skills across the entire practice budget. As shown in Figure 3 , our approach comprises three steps: Competence Prediction. Competence improvement is modeled as a function of the practice budget. This model is initialized with a domain-specific prior and updated online based on the robot’s actual improvement.

Experimental Evaluation

We evaluate our method in three long-horizon table-top manipulation tasks. Our simulated environments are implemented in MuJoCo [ 39 ] using LIBERO [ 27 ] : Cleanup (simulated). A Franka Panda robot must clear a table by placing an object into one of three drawers. The task MDP has $47$ abstract states and $10$ skills in total, with the longest task plan consisting of $4$ skills.

The robot is provided abstract specifications of Pick , Place , OpenDrawer , and CloseDrawer skills. Cleanup-Multi (simulated).

Experimental Evaluation — Deliberate Practice: Learning Robot Skills under a Budget — Figure 4: (a) Breakfast task requires learning forceful manipulation skills to interact with novel articulated objects. (b) Cleanup task requires the robot to place objects in the top, middle, or bottom drawer with rewards $1$, $2$ and $4$ respectively.

Sources and Demonstrations

Figures are reproduced from Vats et al. See the full paper for experimental details and the project page, when the paper links one, for demonstrations and videos.