robotics system builders

Abstract: Goal reasoning (GR) is the process by which intelligent agents continually reason about the goals they are pursuing, which may lead to goal change, for instance when opportunities arise, or when events cause discrepancies among expected and actual outcomes of actions planned to achieve goals. In this paper we investigate whether Hierarchical Task Network (HTN) planning can provide a uniform implementation of the Goal Lifecycle, a flexible process model for goal reasoning in dynamic environments. The Goal Lifecycle progressively refines goals by adding constraints to them through a series of strategies (e.g., {\it formulate, select, expand}) and uses other strategies (e.g., {\it defer, re-expand, repair}) to resolve the current goal when an event impacts it. We present a novel GR agent architecture that employs HTN planning in this Goal Lifecycle. In addition to using HTN planning for the \textit{expand} and \textit{re-expand} strategies, as is frequently done in GR and planning research, we also use it in the \textit{formulate} strategy, to obtain a uniform planning approach that provides several benefits (e.g., the possibility to formulate new goals in view of achieving higher level overarching goals). We also explain how we use an enhanced HTN action representation to process discrepancies among expected and actual action outcomes that can trigger the transition from one goal to another. Finally, we empirically demonstrate that our GR architecture enables simulated autonomous vehicles to adaptively conduct missions in dynamic environments.

Abstract: Mainstream point cloud completion methods are unanimously based on deep learning architectures. On one hand, these methods tend to repetitively stacking attention modules to enhance network performance, which drifts away from the neat design of unified and interpret-able networks; on the other hand, nearly all existing architectures rely solely on feed-forward connections to propagate information, neglecting feedback from deeper layers. The recurrent neural architecture, formed by feedforward-feedback loops, have shown its ability to generate Illusory Contours (ICs) in the V1 brain region of mammals; and ICs are essentially a completion mechanism of the brain for visual information. Motivated by this, we design a Brain-inspired Recurrent Illusory Contour Encoder (BRICE) to "imagine" what is missing in the incomplete 3D point clouds. Building upon the foundation model of Predictive Coding (PC), the plug-and-play nature enables BRICE to be seamlessly integrated into mainstream point cloud completion baselines. Experimental results show that BRICE can enhance various completion baselines on different 3D datasets such as MVP2K, PCN, and PlantPCom. Our code: Https://github.com/wangyunlei-wyl/BRICE.

Abstract: Movable antennas (MAs) have recently emerged as a promising technology to enhance the wireless sensing performance by exploiting additional spatial degrees of freedom. In this work, we consider an MA-enabled near-field target localization problem under a device-based sensing architecture, where the sensing target actively transmits signals and the sensing receiver estimates its location based on the received signals. We adopt the squared position error bound (SPEB) as the sensing performance metric and aim to minimize the target's SPEB by optimizing the antenna placement at the sensing receiver. The resulting optimization problem is highly non-convex due to the minimum antenna spacing constraints and the intricate dependence of the SPEB on the antenna locations and the intractable form of SPEB. To tackle this problem, we first relax the inter-spacing constraints and accordingly formulate an antenna position distribution problem in terms of degree-4 truncated moments. The reformulated moment-based problem admits an equivalent semidefinite programming formulation and can be optimally solved using standard convex optimization solvers. Using truncated moment theory, we then recover an optimal antenna position distribution supported on a finite number of locations and construct an efficient clustered placement policy. Numerical results demonstrate that the constructed optimal distribution exhibits a three-cluster structure with source-dependent weights. The proposed antenna placement method outperforms conventional array geometries and closely approaches the exhaustive-search benchmark scheme.

Abstract: Work on memory for multimodal agents optimizes what is written, updated and retrieved. Between retrieval and the answer, however, is a stage that multimodal memory evaluations do not isolate: what of the retrieved memory reaches the model, and in what form. We call it delivery, and a controlled decomposition on MemLens locates the remaining room there. With the retrieved evidence set exactly fixed, delivering the original pixels instead of withholding them raises accuracy by 13.87 points on an 8B backbone, whereas making retrieval perfect on those same messages improves it by 2.31. Delivery is the larger term on all three MemLens backbones and grows with backbone strength; retrieval grows too, without closing the gap. We propose DeliverMem, an instantiation of delivery as three decisions: keep the original modality, give each item a readable identity, and state when it was seen, with a retrieval-side adapter for the one property delivery cannot supply. Each is measured against a delivery-matched control that alters only its own variable. DeliverMem leads the strongest published memory agent on MemLens at all four context lengths, and beats DMV-Bench's own strongest method at every setting on both backbones. On MemLens it does this on a tenth to a seventieth of the input. Each decision helps only where the question lacks what it supplies, and is null elsewhere. A single fixed configuration nonetheless leads both benchmarks, without training any component or modifying the stored records. Project page: https://avalon-s.github.io/DeliverMem/

Abstract: Load-adaptive gravity balancing mechanisms (LA-GBMs) can accommodate various loading conditions by passively changing their characteristics in response to payload variations. However, their design is difficult because both the desired mechanism motion and static equilibrium under variable payloads must be satisfied simultaneously. This study proposes a general design methodology for LA-GBMs that does not depend on specific mechanism architectures or mechanical elements. The necessary conditions for the potential fields of LA-GBMs are formulated, and two general forms are derived: an affine form representing the effect of payload mass and a factorized form representing state transitions associated with load adaptation and gravity balancing. These forms are then provided to generative AI as design requirements to generate candidate potential functions. The generated functions are analytically verified in terms of their conformity to the two general forms and the conditions required for valid LA-GBMs. Furthermore, the obtained potential functions are decomposed into individual terms, and an example of a method for constructing an LA-GBM by combining springs, counterweights, and function-generating linkage mechanisms is presented. By using potential functions as an intermediate representation, the proposed framework enables the generation of LA-GBM design candidates without prescribing a mechanism architecture in advance. Mechanical realizability and manufacturability of the generated potential fields remain important issues for future work.

Abstract: Vision-language agents that crop and zoom are trained with rewards that credit a successful tool call, yet a successful call does not show that the model needed to look or used the pixels it received. On our cold-start checkpoint only 10% to 12% of visual calls were both needed and used, and released agents make spurious calls 36% to 87% of the time on individual benchmarks. Outcome rewards, judge rewards, and branch probes each observe one side of this failure, and about two thirds of what an outcome reward pays goes to calls that were neither needed nor used. CounterCredit asks both questions of every image-returning call at its realized pre-call state, using the policy's own gold-answer score. A decision value compares the realized visual branch with answering immediately; an evidence value compares the returned crop with random same-size patches substituted into the same call. A call verified on both earns cashback and every other executed call pays rent; the price is bounded so that every correct trajectory outranks every wrong one, and a dual-channel GRPO advantage keeps the price in its own units. From the same cold start, prompt pool, and budget, CounterCredit reaches 89.5% on V*, 80.2% on HR-Bench-4K, and 76.4% on HR-Bench-8K, 6.3 to 9.4 points above outcome-only GRPO at 1.78 against 1.84 calls per question, and lowers the spurious-call rate to 31% to 36%, the lowest among the agents evaluated. The same recipe lifts a Qwen3-VL-8B base from 75.4 to 80.8 on average.

Abstract: Traditional software delivery assumes a static paradigm: code is constructed prior to execution and deployed as a fixed artifact. We present Agentic Just-In-Time Software Construction (A-JIT), a paradigm that replaces static binaries with dynamic, software systems that can perpetually evolve to meet changing demands. In A-JIT, an application is an integrated assembly comprising code, a runtime harness, and an embedded AI agent that continuously observes system usage and live execution traces. Much like a traditional JIT compiler specializes machine code to runtime execution paths, A-JIT specializes software logic, workflows, and tool interfaces to meet the specific needs of the end-user. By integrating synthesis directly into the ambient application lifecycle, A-JIT enables applications to dynamically construct missing implementations, generate new capabilities on the fly, and continuously adapt to end-user behavior. We demonstrate how this model supports trace-driven human-AI co-construction and opens a new design space for adaptive, self-evolving software.