arXiv subjects

Michael Amir

Publications and source records attributed to Michael Amir.

At least 19 recordsLinked to original sources

arXiv2026

Leave It Better Than Found: Value-Guided Cooperation in Sequential Tasks

We consider the problem of sequential task continuation, where one robot acts in a workspace that has been affected by a previous robot. Current approaches in multi-robot cooperation rely heavily on co-training or co-optimization, assumingthat the robots have access to collective interaction data during the policy training or planning process. However, as increasingly varied robotic embodiments and policies are deployed and operate in the same real-world environment, learning policies for cooperation becomes infeasible due to the combinatorial explosion of possible robot combinations that need to be accounted for. To address this challenge, this paper considers the problem of ad hoc sequential cooperation between robots that have never before operated together in a shared environment. We demonstrate that robots can explicitly communicate their environmental preferences through sharing their value functions, enabling coordination without the need for prior joint training. To ground this exchange and prevent overconfidence in out-of-distribution states, we further introduce a k-nearest neighbors (KNN)- based uncertainty filter. Our method yields an increase of up to 48.5 pp compared with naive or even selfish agents, in task success rates across simulated and real-world experiments, demonstrating that value function transmission between agents is useful not merely as an internal learning tool, but as an effective basis for ad hoc coordination. Website: https://proroklab.github.io/PAVE/

cs.RO↗

arXiv2026

Robot Motion as a Keyed Communication Channel

Messages from electronic devices are conventionally received as text, audio, or radio signals. But robots move with rich, articulate motion in the real world, opening up the possibility of transmitting messages through motion itself. In this paper, we consider the problem of motion-based communication, where we seek to modify a robot's movements so as to transmit messages detectable from remote sensing (e.g., video or motion capture), without degrading policy performance. We introduce a method for messaging through motion capable of encoding arbitrary message content over short payloads - such as an agent's current intent - as noise in any pre-trained policy's actions. This brings a new kind of robustness to robot communication: this 'physical' channel complements standard wireless communications channels but does not depend on them, requiring no extra hardware nor the establishment of a direct link to the robot. We systematically characterize the space of encoding schemes and derive design heuristics, then validate them across simulated environments and real-robot deployment; on real robots running at 50 Hz, four robots jointly recover an 8-bit message at an aggregate 0.8 bits/s.

cs.RO↗

arXiv2026

Learning Robot Behavior Priors from Non-Robotic Synthetic Data

Real robot data, rich across tasks and morphologies, is often treated as the bottleneck for training robot learning models, including embodied foundation models that aim for zero-shot physical transfer. In this paper, we investigate the degree to which such models can acquire necessary structure using data from systems without embodiment. To this end, we test if inexpensive, procedurally generated synthetic data can yield useful priors before a model observes any real-world behavior. We study this through robot action prediction, a natural pretraining objective and a direct test of whether structure learned from arbitrary synthetic tasks transfers to embodied settings. We introduce three data generation pipelines that model mathematical abstractions of robotic control, and use them as pretraining priors for a Transformer. We demonstrate that a model trained purely on this non-robotic data attains non-trivial zero-shot prediction performance on unseen physical robot tasks via an in-context learning mechanism, and that our specialized control priors notably improve upon generic causal priors. We also compare against a compute-matched baseline trained on real robot data, and find that it outperforms our model but only on the datasets it was trained on. These findings suggest that procedurally generated, mathematical abstractions can yield useful robotic priors, offering a potential complement to expensive embodied data collection.

cs.RO↗

arXiv2026

Learning Robot Metareasoning: Adaptive Computation Budgets for Real-Time Control

Planning while moving faces an historical trade-off between deliberation and reactivity. A planner running concurrently with execution needs time to compute, so the robot must commit to an open-loop action chunk for that duration: a longer commitment buys a better plan at the cost of feedback frequency. Existing controllers fix this action commitment, and consequently the computation budget, before deployment. However, the optimal computation time is highly state-dependent (due to factors such as unmodeled dynamics, safety constraints, and marginal optimisation gains), i.e. dynamic, therefore offline static tuning risks leaving systems blind during critical maneuvers or unable to find global effective solutions. We cast the identification of the optimal computation budget as a semi-Markov decision process (SMDP) in which a single decision defines both the robot action commitments and the time allocated to the solver. To solve this, we train a reinforcement-learning policy conditioned on the current state, the executing plan, and inexpensive planner statistics. Across ground-robot navigation, agile quadrotor flight, and contact-rich manipulation, the learned policy changes its budget with the demands of the scenario, discovering efficient reasoning patterns that outperform fixed-budget baselines. https://lorenzomagnino.github.io/learning-robot-metareasoning-review/

cs.RO↗

arXiv2026

GAMBIT: Learning to Plan Continuous Multi-Robot Trajectories

GAMBIT is an opening chess move in which a player sacrifices a piece, typically a pawn, to gain a positional advantage later in the game. Analogously, in multi-robot coordination, individual robots may need to forgo locally reward-maximising behaviours to improve overall team performance. Such self-sacrificial behaviours are difficult to capture with manually designed heuristics, particularly in dense, interaction-rich environments. Focusing on double-integrator continuous dynamics, this work studies how to learn such coordinated heuristics over motion primitives for multi-robot trajectory execution. Our framework, GAMBIT, first learns coordinated motion-primitive selection through imitation learning and subsequently fine-tunes the policy through reinforcement learning. We further introduce a safeguarded rollout mechanism with backup trajectories that guarantees collision-free execution at all times. Experiments demonstrate that GAMBIT substantially outperforms a range of baselines, including centralised motion planners and decentralised reactive planners, while exhibiting strong scalability. In particular, it coordinates over a thousand robots with planning latency below a few hundred milliseconds in continuous domains.

cs.RO↗

arXiv2026

Games Over Observation Spaces in Multi-Agent Capture the Flag

We consider a multi-agent Capture the Flag (CtF) scenario in a graph-based environment, where a team of attackers seeks to reach designated flag vertices while a defending team attempts to intercept them. In our setting, both teams operate using decentralized heuristic policies. While the attacking team may choose its heuristic from a diverse library of policies, the defense is restricted to playing a single fixed policy. To overcome this limitation, a centralized defense oracle strategically restricts the portion of the graph visible to each of its agents in order to elicit a wider range of emergent behaviors from its fixed policy. We formalize this interaction as a two-player zero-sum game, where the attacker reasons over its library of heuristics and the defense reasons over the combinatorial space of visibility profiles. To solve this intractably large game, we propose a Double Oracle algorithm to find approximate empirical equilibria, and we empirically validate that observation manipulation can improve the defense's performance.

cs.GT↗

arXiv2026

Open Problems in Constitutional Preference Reconstruction

Pairwise preference data is widely used for training and evaluating language models (e.g., RLHF), but each datapoint records a \emph{choice}, not the rationale behind it. Methods such as Inverse Constitutional AI (ICAI) attempt to improve interpretability by compressing datasets into short ``constitutions'' of natural-language principles. We argue this framing is under-specified: a flat list of principles is not yet an executable decision rule because it leaves principle composition implicit. We use the pairwise setting as a testbed to empirically characterize three open problems in constitutional methods. First, principle quality is hard to measure: coverage and accuracy are useful but incomplete proxies for end-to-end reconstruction. Second, \emph{composition is ambiguous}: holding principles fixed, different executors (LLM judge versus majority vote) agree only $73\%$ of the time. Third, \emph{constitutions differ between LLMs}: cross-model vote agreement is $73\%$, whereas intra-model agreement is $81\%$. Across PRISM, AlpacaEval, and Chatbot Arena, we show that principle refinement (ICAI+) may be a first step towards ameliorating these problems: inter-executor agreement rises to $78\%$, and transparent executors match LLM judge accuracy ($66\%$ vs.\ $67\%$). Our results highlight that constitutions should be evaluated as \emph{constitution--executor systems}, with implications for LLMs-as-a-judge broadly.

cs.AI↗

arXiv2026

Scaling Multi-Agent Environment Co-Design with Diffusion Models

The agent-environment co-design paradigm jointly optimises agent policies and environment configurations in search of improved system performance. With application domains ranging from warehouse logistics to windfarm management, co-design promises to fundamentally change how we deploy multi-agent systems. However, current co-design methods struggle to scale. They collapse under high-dimensional environment design spaces and suffer from sample inefficiency when addressing moving targets inherent to joint optimisation. We address these challenges by developing Diffusion Co-Design (DiCoDe), a scalable and sample-efficient co-design framework pushing co-design towards practically relevant settings. DiCoDe incorporates two core innovations. First, we introduce Projected Universal Guidance (PUG), a sampling technique that enables DiCoDe to explore a distribution of reward-maximising environments while satisfying hard constraints such as spatial separation between obstacles. Second, we devise a critic distillation mechanism to share knowledge from the reinforcement learning critic, ensuring that the guided diffusion model adapts to evolving agent policies using a dense and up-to-date learning signal. Together, these improvements lead to superior environment-policy pairs when validated on challenging multi-agent environment co-design benchmarks including warehouse automation, multi-agent pathfinding and wind farm optimisation. Our method consistently exceeds the state-of-the-art, achieving, for example, 39% higher rewards in the warehouse setting with 66% fewer simulation samples. This sets a new standard in agent-environment co-design, and is a stepping stone towards reaping the rewards of co-design in real world domains.

cs.LG↗

arXiv2026

Pairwise is Not Enough: Hypergraph Neural Networks for Multi-Agent Pathfinding

Multi-Agent Path Finding (MAPF) is a representative multi-agent coordination problem, where multiple agents are required to navigate to their respective goals without collisions. Solving MAPF optimally is known to be NP-hard, leading to the adoption of learning-based approaches to alleviate the online computational burden. Prevailing approaches, such as Graph Neural Networks (GNNs), are typically constrained to pairwise message passing between agents. However, this limitation leads to suboptimal behaviours and critical issues, such as attention dilution, particularly in dense environments where group (i.e. beyond just two agents) coordination is most critical. Despite the importance of such higher-order interactions, existing approaches have not been able to fully explore them. To address this representational bottleneck, we introduce HMAGAT (Hypergraph Multi-Agent Attention Network), a novel architecture that leverages attentional mechanisms over directed hypergraphs to explicitly capture group dynamics. Empirically, HMAGAT establishes a new state-of-the-art among learning-based MAPF solvers: e.g., despite having just 1M parameters and being trained on 100$\times$ less data, it outperforms the current SoTA 85M parameter model. Through detailed analysis of HMAGAT's attention values, we demonstrate how hypergraph representations mitigate the attention dilution inherent in GNNs and capture complex interactions where pairwise methods fail. Our results illustrate that appropriate inductive biases are often more critical than the training data size or sheer parameter count for multi-agent problems.

cs.LG↗

arXiv2026

When Is Diversity Rewarded in Cooperative Multi-Agent Learning?

The success of teams in robotics, nature, and society often depends on the division of labor among diverse specialists; however, a principled explanation for when such diversity surpasses a homogeneous team is still missing. Focusing on multi-agent task allocation problems, we study this question from the perspective of reward design: what kinds of objectives are best suited for heterogeneous teams? We first consider an instantaneous, non-spatial setting where the global reward is built by two generalized aggregation operators: an inner operator that maps the $N$ agents' effort allocations on individual tasks to a task score, and an outer operator that merges the $M$ task scores into the global team reward. We prove that the curvature of these operators determines whether heterogeneity can increase reward, and that for broad reward families this collapses to a simple convexity test. Next, we ask what incentivizes heterogeneity to emerge when embodied, time-extended agents must learn an effort allocation policy. To study heterogeneity in such settings, we use multi-agent reinforcement learning (MARL) as our computational paradigm, and introduce Heterogeneity Gain Parameter Search (HetGPS), a gradient-based algorithm that optimizes the parameter space of underspecified MARL environments to find scenarios where heterogeneity is advantageous. Across different environments, we show that HetGPS rediscovers the reward regimes predicted by our theory to maximize the advantage of heterogeneity, both validating HetGPS and connecting our theoretical insights to reward design in MARL. Together, these results help us understand when behavioral diversity delivers a measurable benefit.

cs.MA↗

arXiv2025

Remotely Detectable Robot Policy Watermarking

The success of machine learning for real-world robotic systems has created a new form of intellectual property: the trained policy. This raises a critical need for novel methods that verify ownership and detect unauthorized, possibly unsafe misuse. While watermarking is established in other domains, physical policies present a unique challenge: remote detection. Existing methods assume access to the robot's internal state, but auditors are often limited to external observations (e.g., video footage). This ``Physical Observation Gap'' means the watermark must be detected from signals that are noisy, asynchronous, and filtered by unknown system dynamics. We formalize this challenge using the concept of a \textit{glimpse sequence}, and introduce Colored Noise Coherency (CoNoCo), the first watermarking strategy designed for remote detection. CoNoCo embeds a spectral signal into the robot's motions by leveraging the policy's inherent stochasticity. To show it does not degrade performance, we prove CoNoCo preserves the marginal action distribution. Our experiments demonstrate strong, robust detection across various remote modalities, including motion capture and side-way/top-down video footage, in both simulated and real-world robot experiments. This work provides a necessary step toward protecting intellectual property in robotics, offering the first method for validating the provenance of physical policies non-invasively, using purely remote observations.

cs.RO↗

arXiv2025

Graph Attention-Guided Search for Dense Multi-Agent Pathfinding

Finding near-optimal solutions for dense multi-agent pathfinding (MAPF) problems in real-time remains challenging even for state-of-the-art planners. To this end, we develop a hybrid framework that integrates a learned heuristic derived from MAGAT, a neural MAPF policy with a graph attention scheme, into a leading search-based algorithm, LaCAM. While prior work has explored learning-guided search in MAPF, such methods have historically underperformed. In contrast, our approach, termed LaGAT, outperforms both purely search-based and purely learning-based methods in dense scenarios. This is achieved through an enhanced MAGAT architecture, a pre-train-then-fine-tune strategy on maps of interest, and a deadlock detection scheme to account for imperfect neural guidance. Our results demonstrate that, when carefully designed, hybrid search offers a powerful solution for tightly coupled, challenging multi-agent coordination problems.

cs.AI↗

arXiv2025

ReCoDe: Reinforcement Learning-based Dynamic Constraint Design for Multi-Agent Coordination

Constraint-based optimization is a cornerstone of robotics, enabling the design of controllers that reliably encode task and safety requirements such as collision avoidance or formation adherence. However, handcrafted constraints can fail in multi-agent settings that demand complex coordination. We introduce ReCoDe--Reinforcement-based Constraint Design--a decentralized, hybrid framework that merges the reliability of optimization-based controllers with the adaptability of multi-agent reinforcement learning. Rather than discarding expert controllers, ReCoDe improves them by learning additional, dynamic constraints that capture subtler behaviors, for example, by constraining agent movements to prevent congestion in cluttered scenarios. Through local communication, agents collectively constrain their allowed actions to coordinate more effectively under changing conditions. In this work, we focus on applications of ReCoDe to multi-agent navigation tasks requiring intricate, context-based movements and consensus, where we show that it outperforms purely handcrafted controllers, other hybrid approaches, and standard MARL baselines. We give empirical (real robot) and theoretical evidence that retaining a user-defined controller, even when it is imperfect, is more efficient than learning from scratch, especially because ReCoDe can dynamically change the degree to which it relies on this controller.

cs.RO↗

arXiv2025

Time, Travel, and Energy in the Uniform Dispersion Problem

We investigate the algorithmic problem of uniformly dispersing a swarm of robots in an unknown, gridlike environment. In this setting, our goal is to study the relationships between performance metrics and robot capabilities. We introduce a formal model comparing dispersion algorithms based on makespan, traveled distance, energy consumption, sensing, communication, and memory. Using this framework, we classify uniform dispersion algorithms according to their capability requirements and performance. We prove that while makespan and travel can be minimized in all environments, energy cannot, if the swarm's sensing range is bounded. In contrast, we show that energy can be minimized by ``ant-like'' robots in synchronous settings and asymptotically minimized in asynchronous settings, provided the environment is topologically simply connected, by using our ``Find-Corner Depth-First Search'' (FCDFS) algorithm. Our theoretical and experimental results show that FCDFS significantly outperforms known algorithms. Our findings reveal key limitations in designing swarm robotics systems for unknown environments, emphasizing the role of topology in energy-efficient dispersion.

cs.RO↗

arXiv2024

Patrolling Grids with a Bit of Memory

This work addresses the challenge of patrolling regular grid graphs of any dimension using a single mobile agent with minimal memory and limited sensing range. We show that it is impossible to patrol some grid graphs with $0$ bits of memory, regardless of sensing range, and give an exact characterization of those grid graphs that can be patrolled with $0$ bits of memory and sensing range $V$. On the other hand, we show that an algorithm exists using $1$ bit of memory and $V=1$ that patrols any $d$-dimensional grid graph. This result is surprising given that the agent must be able to move in $2d$ distinct directions to patrol, while $1$ bit of memory allows specifying only two directions per sensory input. Our $1$-bit patrolling algorithm handles this by carefully exploiting a small state-space to access all the needed directions while avoiding getting stuck. Overall, our results give concrete evidence that extremely little memory is needed for patrolling highly regular environments like grid graphs compared to arbitrary graphs. The techniques we use, such as partitioning the environment into sensing regions and exploiting distinct coordinates resulting from higher-dimensionality, may be applicable to analyzing the space complexity of patrolling in other types of regular environments as well.

cs.RO↗

arXiv2023

Stigmergy-based, Dual-Layer Coverage of Unknown Indoor Regions

We present algorithms for uniformly covering an unknown indoor region with a swarm of simple, anonymous and autonomous mobile agents. The exploration of such regions is made difficult by the lack of a common global reference frame, severe degradation of radio-frequency communication, and numerous ground obstacles. We propose addressing these challenges by using airborne agents, such as Micro Air Vehicles, in dual capacity, both as mobile explorers and (once they land) as beacons that help other agents navigate the region. The algorithms we propose are designed for a swarm of simple, identical, ant-like agents with local sensing capabilities. The agents enter the region, which is discretized as a graph, over time from one or more entry points and are tasked with occupying all of its vertices. Unlike many works in this area, we consider the requirement of informing an outside operator with limited information that the coverage mission is complete. Even with this additional requirement we show, both through simulations and mathematical proofs, that the dual role concept results in linear-time termination, while also besting many well-known algorithms in the literature in terms of energy use.

cs.MA↗

arXiv2022

Multi-Agent Distributed and Decentralized Geometric Task Allocation

We consider the general problem of geometric task allocation, wherein a large, decentralised swarm of simple mobile agents must detect the locations of tasks in the plane and position themselves nearby. The tasks are represented by an a priori unknown demand profile $Φ(x,y)$ that determines how many agents are needed in each location. The agents are autonomous, oblivious and indistinguishable, and have finite sensing range. They must configure themselves according to $Φ$ using only local information about $Φ$ and about the positions of nearby agents. All agents act according to the same local sensing-based rule of motion, and cannot explicitly communicate nor share information. We propose an optimization-based approach to the problem which results in attraction-repulsion dynamics. Repulsion encourages agents to spread out and explore the region so as to find the tasks, and attraction causes them to accumulate at task locations. We derive this approach via gradient descent over an appropriate ``error'' functional, and test it extensively through numerical simulations. The figures in this work are snapshots of simulations that can be viewed online at https://youtu.be/kyUiGYSaaoQ.

cs.MA↗

arXiv2022

Optimal Physical Sorting of Mobile Agents

Given a collection of red and blue mobile agents located on two grid rows, we seek to move all the blue agents to the far left side and all the red agents to the far right side, thus \textit{physically sorting} them according to color. The agents all start on the bottom row. They move simultaneously at discrete time steps and must not collide. Our goal is to design a centralized algorithm that controls the agents so as to sort them in the least number of time steps. We derive an \textbf{exact} lower bound on the amount of time any algorithm requires to sort a given initial configuration of agents. We find an instance optimal algorithm that provably matches this lower bound, attaining the best possible sorting time for any initial configuration. Surprisingly, we find that whenever the leftmost agent is red and the rightmost agent is blue, a straightforward decentralized and local sensing-based algorithm is at most $1$ time step slower than the centralized instance-optimal algorithm.

cs.DM↗