Gioele Zardini

Yuyang Du, Yujun Huang, Gioele Zardini

Designing a neural network processor is an end-to-end co-design problem: network architecture and training budget determine the inference workload; hardware mapping decisions determine chip area, latency, and energy; and these characteristics govern fabrication yield and manufacturing cost. In practice, these decisions are made in separate stages, and existing co-design methodologies are tightly coupled to specific algorithms, making it difficult to improve one component without reworking the entire pipeline. This paper presents a unified framework, grounded in monotone co-design theory, that composes four interoperable design blocks spanning network training, chip mapping, wafer-level fabrication, and compute resource allocation. Each block exposes only a functionality-resource interface to the rest of the system, so any block can be refined without structural changes elsewhere. A central contribution is the treatment of uncertainty: rather than collapsing stochastic outcomes into point estimates, the framework introduces Confidence, the inverse of success probability, as an explicit and optimizable resource alongside cost, time, and power. Three case studies validate the approach. The first recovers Pareto-optimal implementations across heterogeneous application scenarios. The second confirms that Confidence functions as a continuously tunable design knob rather than a post-hoc diagnostic. The third demonstrates that improving a single block's implementation set automatically propagates to the global Pareto front, without modifying the co-design diagram.

Aditya Gahlawat, Ahmed Aboudonia, Sandeep Banik et al.

Imitation learning (IL) enables autonomous behavior by learning from expert demonstrations. While more sample-efficient than comparative alternatives like reinforcement learning, IL is sensitive to compounding errors induced by distribution shifts. There are two significant sources of distribution shifts when using IL-based feedback laws on systems: distribution shifts caused by policy error and distribution shifts due to exogenous disturbances and endogenous model errors due to lack of learning. Our previously developed approaches, Taylor Series Imitation Learning (TaSIL) and $\mathcal{L}_1$ -Distributionally Robust Adaptive Control (\ellonedrac), address the challenge of distribution shifts in complementary ways. While TaSIL offers robustness against policy error-induced distribution shifts, \ellonedrac offers robustness against distribution shifts due to aleatoric and epistemic uncertainties. To enable certifiable IL for learned and/or uncertain dynamical systems, we formulate \textit{Distributionally Robust Imitation Policy (DRIP)} architecture, a Layered Control Architecture (LCA) that integrates TaSIL and~\ellonedrac. By judiciously designing individual layer-centric input and output requirements, we show how we can guarantee certificates for the entire control pipeline. Our solution paves the path for designing fully certifiable autonomy pipelines, by integrating learning-based components, such as perception, with certifiable model-based decision-making through the proposed LCA approach.

Yujun Huang, Gioele Zardini

Complex engineered systems require coordinated design choices across heterogeneous components under multiple conflicting objectives and uncertain specifications. Monotone co-design provides a compositional framework for such problems by modeling each subsystem as a design problem: a feasible relation between provided functionalities and required resources in partially ordered sets. Existing uncertain co-design models rely on interval bounds, which support worst-case reasoning but cannot represent probabilistic risk or multi-stage adaptive decisions. We develop a distributional extension of co-design that models uncertain design outcomes as distributions over design problems and supports adaptive decision processes through Markov-kernel re-parameterizations. Using quasi-measurable and quasi-universal spaces, we show that the standard co-design interconnection operations remain compositional under this richer notion of uncertainty. We further introduce queries and observations that extract probabilistic design trade-offs, including feasibility probabilities, confidence bounds, and distributions of minimal required resources. A task-driven unmanned aerial vehicle case study illustrates how the framework captures risk-sensitive and information-dependent design choices that interval-based models cannot express.

Mingjia He, Zhiyu He, Jan Ghadamian et al.

Mobility systems are complex socio-technical environments influenced by multiple stakeholders with hierarchically interdependent decisions, rendering effective control and policy design inherently challenging. We bridge hierarchical game-theoretic modeling with online feedback optimization by casting urban mobility as a tri-level Stackelberg game (travelers, operators, municipality) closed in a feedback loop. The municipality iteratively updates taxes, subsidies, and operational constraints using a projected two-point (gradient-free) scheme, while lower levels respond through equilibrium computations (Frank-Wolfe for traveler equilibrium; operator best responses). This model-free pipeline enforces constraints, accommodates heterogeneous users and modes, and scales to higher-dimensional policy vectors without differentiating through equilibrium maps. On a real multimodal network for Zurich, Switzerland, our method attains substantially better municipal objectives than Bayesian optimization and Genetic algorithms, and identifies integration incentives that increase multimodal usage while improving both operator objectives. The results show that feedback-based regulation can steer competition toward cooperative outcomes and deliver tangible welfare gains in complex, data-rich mobility ecosystems.

Jiarui Li, Runyu Zhang, Gioele Zardini

Multi-agent coordination in automated warehouses and logistics is commonly modeled as the Multi-Agent Path Finding (MAPF) problem. Closed-loop MAPF algorithms improve scalability by planning only the next movement and replanning online, but this finite-horizon viewpoint can be shortsighted and makes it difficult to preserve global guarantees and exploit compositional structure. This issue is especially visible in Anytime Closed-Loop Conflict-Based Search (ACCBS), which applies Conflict-Based Search (CBS) over dynamically extended finite horizons but, under finite computational budgets, may terminate with short active prefixes in dense instances. We introduce certificate trajectories and their associated fleet budget as a general mechanism for filtering closed-loop updates. A certificate provides a conflict-free fallback plan and a monotone upper bound on the remaining cost; accepting only certificate-improving updates yields completeness. The same budget information induces a budget-limited factorization that enables global, inheritable decomposition across timesteps. Instantiating the framework on ACCBS yields Certificate-Driven Conflict-Based Search (CDCBS). Experiments on benchmark maps show that CDCBS achieves more consistent solution quality than ACCBS, particularly in dense settings, while the proposed factorization reduces effective group size.

Yubo Cai, Yujun Huang, Meshal Alharbi et al.

Designing complex engineered systems requires managing tightly coupled trade-offs between subsystem capabilities and resource requirements. Monotone co-design provides a compositional language for such problems, but its generality does not by itself reveal which problem classes admit exact and scalable computation. This paper isolates such a class by introducing Linear Design Problems (LDPs): design problems whose feasible functionality--resource relations are polyhedra over Euclidean posets. We show that queries on LDPs reduce exactly to Multi-Objective Linear Programs (MOLPs), thereby connecting monotone co-design semantics with polyhedral multiobjective optimization. We further prove that LDPs are closed under the fundamental co-design interconnections, implying that any interconnection of linear components induces a system-level LDP. To compute the resulting feasible sets, we develop two complementary constructions: a monolithic lifted formulation that preserves block-angular sparsity, and a compositional formulation that incrementally eliminates internal variables through polyhedral projection. Beyond the exact linear setting, we show that convex co-design resource queries admit arbitrarily accurate polyhedral outer approximations, with recession-cone error identically zero for standard nonnegative resource cones. Numerical studies on synthetic series-chain benchmarks, a gripper, and a rover co-design validate the theory.

Luigi Tresca, Carolin Schmidt, James Harrison et al.

Fleets of robo-taxis offering on-demand transportation services, commonly known as Autonomous Mobility-on-Demand (AMoD) systems, hold significant promise for societal benefits, such as reducing pollution, energy consumption, and urban congestion. However, orchestrating these systems at scale remains a critical challenge, with existing coordination algorithms often failing to exploit the systems' full potential. This work introduces a novel decision-making framework that unites mathematical modeling with data-driven techniques. In particular, we present the AMoD coordination problem through the lens of reinforcement learning and propose a graph network-based framework that exploits the main strengths of graph representation learning, reinforcement learning, and classical operations research tools. Extensive evaluations across diverse simulation fidelities and scenarios demonstrate the flexibility of our approach, achieving superior system performance, computational efficiency, and generalizability compared to prior methods. Finally, motivated by the need to democratize research efforts in this area, we release publicly available benchmarks, datasets, and simulators for network-level coordination alongside an open-source codebase designed to provide accessible simulation platforms and establish a standardized validation process for comparing methodologies. Code available at: https://github.com/StanfordASL/RL4AMOD

Johannes Gaber, Meshal Alharbi, Daniele Gammelli et al.

Large robot fleets are now common in warehouses and other logistics settings, where small control gains translate into large operational impacts. In this article, we address task scheduling for lifelong Multi-Agent Pickup-and-Delivery (MAPD) and propose a hybrid method that couples learning-based global guidance with lightweight optimization. A graph neural network policy trained via reinforcement learning outputs a desired distribution of free agents over an aggregated warehouse graph. This signal is converted into region-to-region rebalancing through a minimum-cost flow, and finalized by small, local assignment problems, preserving accuracy while keeping per-step latency within a 1 s compute budget. On congested warehouse benchmarks from the League of Robot Runners (LRR) with up to 500 agents, our approach improves throughput by up to 10% over the 2024 winning scheduler while maintaining real-time execution. The results indicate that coupling graph-structured learned guidance with tractable solvers reduces congestion and yields a practical, scalable blueprint for high-throughput scheduling in large fleets.

Lee Marom, Skylar Tibbits, Gioele Zardini et al.

Natural materials achieve adaptive behavior through hierarchical organization and coupled mechanisms across scales. Their translation into engineering, however, remains largely heuristic. What is missing is a formal translation framework that carries biological design logic into engineered realization while preserving physical consistency across levels of abstraction. Here we present a category theoretic compositional framework for verified nature-derived design. The framework defines a category of stimulus response dynamical systems with natural and artificial subcategories. It introduces a structure preserving implementation functor from biological mechanics to engineered systems. It also formalizes a machine agnostic specification layer that links behavioral intent to executable fabrication programs. We instantiate the framework on the hygromorphic pinecone hierarchy as a representative biological case. We implement the full pipeline in Grasshopper, where formal specifications are translated into modular parametric scripts that preserve the compositional structure of the model. The resulting designs are fabricated by fused filament fabrication, evaluated experimentally, and tested against model predictions derived from the pipeline. The current implementation generates four actuator classes spanning two stimulus types and two kinematic responses. One actuator arises purely through composition from previously validated components, without additional manual derivation. The results show that compositionality can function not just as a descriptive language, but as a generative and system level verifiable method for mechanical material design. More broadly, the work provides a concrete route for embedding formal multiscale reasoning within increasingly computational, generative, and physics-driven design workflows.

Maximilian Stralz, Meshal Alharbi, Yujun Huang et al.

Designing multi-agent robotic systems requires reasoning across tightly coupled decisions spanning heterogeneous domains, including robot design, fleet composition, and planning. Much effort has been devoted to isolated improvements in these domains, whereas system-level co-design considering trade-offs and task requirements remains underexplored. In this work, we present a formal and compositional framework for the task-driven co-design of heterogeneous multi-robot systems. Building on a monotone co-design theory, we introduce general abstractions of robots, fleets, planners, executors, and evaluators as interconnected design problems with well-defined interfaces that are agnostic to both implementations and tasks. This structure enables efficient joint optimization of robot design, fleet composition, and planning under task-specific performance constraints. A series of case studies demonstrates the capabilities of the framework. Various component models can be seamlessly incorporated, including new robot types, task profiles, and probabilistic sensing objectives, while non-obvious design alternatives are systematically uncovered with optimality guarantees. The results highlight the flexibility, scalability, and interpretability of the proposed approach, and illustrate how formal co-design enables principled reasoning about complex heterogeneous multi-robot systems.

Dejan Milojevic, Gioele Zardini, Miriam Elser et al.

This paper discusses the integration challenges and strategies for designing mobile robots, by focusing on the task-driven, optimal selection of hardware and software to balance safety, efficiency, and minimal usage of resources such as costs, energy, computational requirements, and weight. We emphasize the interplay between perception and motion planning in decision-making by introducing the concept of occupancy queries to quantify the perception requirements for sampling-based motion planners. Sensor and algorithm performance are evaluated using False Negative Rates (FPR) and False Positive Rates (FPR) across various factors such as geometric relationships, object properties, sensor resolution, and environmental conditions. By integrating perception requirements with perception performance, an Integer Linear Programming (ILP) approach is proposed for efficient sensor and algorithm selection and placement. This forms the basis for a co-design optimization that includes the robot body, motion planner, perception pipeline, and computing unit. We refer to this framework for solving the co-design problem of mobile robots as CODEI, short for Co-design of Embodied Intelligence. A case study on developing an Autonomous Vehicle (AV) for urban scenarios provides actionable information for designers, and shows that complex tasks escalate resource demands, with task performance affecting choices of the autonomy stack. The study demonstrates that resource prioritization influences sensor choice: cameras are preferred for cost-effective and lightweight designs, while lidar sensors are chosen for better energy and computational efficiency.

Meshal Alharbi, Munther A. Dahleh, Gioele Zardini

Many engineered systems must balance competing objectives, such as performance and safety, cost and reliability, or efficiency and sustainability, and are naturally modeled as compositions of interacting subsystems. We study online multi-objective decision-making in monotone co-design, where functionalities and resources are partially ordered, and the goal is to identify the target-feasible antichain of non-dominated trade-offs using few expensive evaluations. We introduce optimistic evaluators: history-dependent bounds on functionality and resource mappings that enable safe elimination of implementations before full evaluation. Based on these evaluators, we develop an elimination-based rejection-sampling algorithm, prove its soundness, and show that the admissible region shrinks monotonically as information accumulates. We instantiate the framework under monotonicity, Lipschitz continuity, and linear-parametric structure. For compositional co-design problems modeled by multigraphs, we show how local optimistic certificates propagate through the tractable remainder of the graph to yield system-level optimistic feasibility and resource bounds. Experiments on multi-robot fleet design, intermodal mobility systems, and synthetic monotone and Lipschitz benchmarks show substantial sample-efficiency gains over uniform sampling, Bayesian optimization, and multi-objective evolutionary algorithms.

Davide Celestini, Amirhossein Afsharrad, Daniele Gammelli et al. · stanford

Effective trajectory generation is essential for reliable on-board spacecraft autonomy. Among other approaches, learning-based warm-starting represents an appealing paradigm for solving the trajectory generation problem, effectively combining the benefits of optimization- and data-driven methods. Current approaches for learning-based trajectory generation often focus on fixed, single-scenario environments, where key scene characteristics, such as obstacle positions or final-time requirements, remain constant across problem instances. However, practical trajectory generation requires the scenario to be frequently reconfigured, making the single-scenario approach a potentially impractical solution. To address this challenge, we present a novel trajectory generation framework that generalizes across diverse problem configurations, by leveraging high-capacity transformer neural networks capable of learning from multimodal data sources. Specifically, our approach integrates transformer-based neural network models into the trajectory optimization process, encoding both scene-level information (e.g., obstacle locations, initial and goal states) and trajectory-level constraints (e.g., time bounds, fuel consumption targets) via multimodal representations. The transformer network then generates near-optimal initial guesses for non-convex optimization problems, significantly enhancing convergence speed and performance. The framework is validated through extensive simulations and real-world experiments on a free-flyer platform, achieving up to 30% cost improvement and 80% reduction in infeasible cases with respect to traditional approaches, and demonstrating robust generalization across diverse scenario variations.

Vincent Abbott, Gioele Zardini

Optimizing deep learning algorithms currently requires slow, manual derivation, potentially leaving much performance untapped. Methods like FlashAttention have achieved a x6 performance improvement over native PyTorch by avoiding unnecessary data transfers, but required three iterations over three years to be developed. Automated compiled methods have consistently lagged behind. This paper extends Neural Circuit Diagrams for deep learning models to consider resource usage and the distribution of tasks across a GPU hierarchy. We show how diagrams can use simple relabellings to derive high-level streaming and tiling optimization strategies along with performance models. We show how this high-level performance model allows the effects of quantization and multi-level GPU hierarchies to be readily considered. We develop a methodology for representing intermediate-level pseudocode with diagrams, allowing hardware-aware algorithms to be derived step-by-step. Finally, we show how our methodology can be used to better understand existing techniques like FlashAttention. This work uses a theoretical framework to link assumptions about GPU behaviour to claims about performance. We aim to lay the groundwork for a scientific approach to GPU optimization where experiments can address clear hypotheses rather than post-hoc rationalizations.

Vincent Abbott, Gioele Zardini

Despite deep learning models running well-defined mathematical functions, we lack a formal mathematical framework for describing model architectures. Ad-hoc notation, diagrams, and pseudocode poorly handle nonlinear broadcasting and the relationship between individual components and composed models. This paper introduces a categorical framework for deep learning models that formalizes broadcasting through the novel axis-stride and array-broadcasted categories. This allows the mathematical function underlying architectures to be precisely expressed and manipulated in a compositional manner. These mathematical definitions are translated into human manageable diagrams and machine manageable data structures. We provide a mirrored implementation in Python (pyncd) and TypeScript (tsncd) to show the universal aspect of our framework, along with features including algebraic construction, graph conversion, PyTorch compilation and diagram rendering. This lays the foundation for a systematic, formal approach to deep learning model design and analysis.

Avalpreet Singh Brar, Rong Su, Jaskaranveer Kaur et al.

Human driver participation is a critical source of uncertainty in Mobility-on-Demand (MoD) rebalancing. Drivers follow platform recommendations probabilistically, and their willingness to comply evolves with experienced outcomes. This creates a closed-loop feedback in which stronger recommendations increase participation, participation increases congestion, congestion lowers allocation success, and realized allocations update adherence beliefs. We propose a microscopic stochastic model that couples (i) belief-driven participation, (ii) Poisson demand, (iii) uniform matching, and (iv) Beta--Bernoulli belief updates. Under a large-population closure, we derive a deterministic mean-field recursion for the population adherence state under platform actuation. For i.i.d. Poisson demand and constant recommendation intensity, we prove global well-posedness and invariance of the recursion, establish equilibrium existence, provide uniqueness conditions, and show global convergence in the regime where platform recommendations are no weaker than baseline participation. We then define steady-state adherence and throughput, characterize the induced performance frontier, and show that adherence and throughput cannot, in general, be simultaneously maximized under uniform time-invariant actuation. This yields a throughput-maximization problem with an adherence floor. Exploiting the monotone frontier structure, we show the optimal uniform time-invariant policy is the maximal feasible recommendation intensity and provide an efficient bisection-based algorithm.

Hans Riess, Yujun Huang, Matthew Klawonn et al.

Monotone co-design enables compositional engineering design by modeling components through feasibility relations between required resources and provided functionalities. However, its standard boolean formulation cannot natively represent quantitative criteria such as cost, confidence, or implementation choice. In practice, these quantities are often introduced through ad hoc scalarization or by augmenting the resource space, which obscures system structure and increases computational burden. We address this limitation by developing a quantale-enriched theory of co-design. We model resources and functionalities as quantale-enriched categories and design problems as quantale-enriched profunctors, thereby lifting co-design from boolean feasibility to general quantitative evaluation. We show that the fundamental operations of series, parallel, and feedback composition remain valid over arbitrary commutative quantales. We further introduce heterogeneous composition through change-of-base maps between quantales, enabling different subsystems to be evaluated in different local semantics and then composed in a common framework. The resulting theory unifies feasibility-, cost-, confidence-, and implementation-aware co-design within one compositional formalism. Numerical examples on a target-tracking system and a UAV delivery problem demonstrate the framework and highlight how native quantitative enrichment can avoid the architectural and computational drawbacks of boolean-only formulations.

Runyu Zhang, Na Li, Asuman Ozdaglar et al.

Risk sensitivity has become a central theme in reinforcement learning (RL), where convex risk measures and robust formulations provide principled ways to model preferences beyond expected return. Recent extensions to multi-agent RL (MARL) have largely emphasized the risk-averse setting, prioritizing robustness to uncertainty. In cooperative MARL, however, such conservatism often leads to suboptimal equilibria, and a parallel line of work has shown that optimism can promote cooperation. Existing optimistic methods, though effective in practice, are typically heuristic and lack theoretical grounding. Building on the dual representation for convex risk measures, we propose a principled framework that interprets risk-seeking objectives as optimism. We introduce optimistic value functions, which formalize optimism as divergence-penalized risk-seeking evaluations. Building on this foundation, we derive a policy-gradient theorem for optimistic value functions, including explicit formulas for the entropic risk/KL-penalty setting, and develop decentralized optimistic actor-critic algorithms that implement these updates. Empirical results on cooperative benchmarks demonstrate that risk-seeking optimism consistently improves coordination over both risk-neutral baselines and heuristic optimistic methods. Our framework thus unifies risk-sensitive learning and optimism, offering a theoretically grounded and practically effective approach to cooperation in MARL.

Vincent Abbott, Kotaro Kamiya, Gerard Glowacki et al.

How do we enable artificial intelligence models to improve themselves? This is central to exponentially improving generalized artificial intelligence models, which can improve their own architecture to handle new problem domains in an efficient manner that leverages the latest hardware. However, current automated compilation methods are poor, and efficient algorithms require years of human development. In this paper, we use neural circuit diagrams, based in category theory, to prove a general theorem related to deep learning algorithms, guide the development of a novel attention algorithm catered to the domain of gene regulatory networks, and produce a corresponding efficient kernel. The algorithm we propose, spherical attention, shows that neural circuit diagrams enable a principled and systematic method for reasoning about deep learning architectures and providing high-performance code. By replacing SoftMax with an $L^2$ norm as suggested by diagrams, it overcomes the special function unit bottleneck of standard attention while retaining the streaming property essential to high-performance. Our diagrammatically derived \textit{FlashSign} kernel achieves comparable performance to the state-of-the-art, fine-tuned FlashAttention algorithm on an A100, and $3.6\times$ the performance of PyTorch. Overall, this investigation shows neural circuit diagrams' suitability as a high-level framework for the automated development of efficient, novel artificial intelligence architectures.

Alessandro Zanardi, Gioele Zardini, Sirish Srinivasan et al.

Modern applications require robots to comply with multiple, often conflicting rules and to interact with the other agents. We present Posetal Games as a class of games in which each player expresses a preference over the outcomes via a partially ordered set of metrics. This allows one to combine hierarchical priorities of each player with the interactive nature of the environment. By contextualizing standard game theoretical notions, we provide two sufficient conditions on the preference of the players to prove existence of pure Nash Equilibria in finite action sets. Moreover, we define formal operations on the preference structures and link them to a refinement of the game solutions, showing how the set of equilibria can be systematically shrunk. The presented results are showcased in a driving game where autonomous vehicles select from a finite set of trajectories. The results demonstrate the interpretability of results in terms of minimum-rank-violation for each player.

Gioele Zardini, David I. Spivak, Andrea Censi et al.

A compositional sheaf-theoretic framework for the modeling of complex event-based systems is presented. We show that event-based systems are machines, with inputs and outputs, and that they can be composed with machines of different types, all within a unified, sheaf-theoretic formalism. We take robotic systems as an exemplar of complex systems and rigorously describe actuators, sensors, and algorithms using this framework.

Gioele Zardini, Andrea Censi, Emilio Frazzoli

Designing cyber-physical systems is a complex task which requires insights at multiple abstraction levels. The choices of single components are deeply interconnected and need to be jointly studied. In this work, we consider the problem of co-designing the control algorithm as well as the platform around it. In particular, we leverage a monotone theory of co-design to formalize variations of the LQG control problem as monotone feasibility relations. We then show how this enables the embedding of control co-design problems in the higher level co-design problem of a robotic platform. We illustrate the properties of our formalization by analyzing the co-design of an autonomous drone performing search-and-rescue tasks and show how, given a set of desired robot behaviors, we can compute Pareto efficient design solutions.

Gioele Zardini, Dejan Milojevic, Andrea Censi et al.

We consider the problem of co-designing embodied intelligence as a whole in a structured way, from hardware components such as propulsion systems and sensors to software modules such as control and perception pipelines. We propose a principled approach to formulate and solve complex embodied intelligence co-design problems, leveraging a monotone co-design theory. The methods we propose are intuitive and integrate heterogeneous engineering disciplines, allowing analytical and simulation-based modeling techniques and enabling interdisciplinarity. We illustrate through a case study how, given a set of desired behaviors, our framework is able to compute Pareto efficient solutions for the entire hardware and software stack of a self-driving vehicle.

Gioele Zardini, David I. Spivak, Andrea Censi et al.

A compositional sheaf-theoretic framework for the modeling of complex event-based systems is presented. We show that event-based systems are machines, with inputs and outputs, and that they can be composed with machines of different types, all within a unified, sheaf-theoretic formalism. We take robotic systems as an exemplar of complex systems and rigorously describe actuators, sensors, and algorithms using this framework.