Recent advances in large language models (LLMs) and autonomous agents have enabled systems capable of performing complex tasks across domains such as human-computer interaction, planning, and web navigation. However, many existing frameworks struggle in real-world or resource-constrained environments due to their reliance on cloud-based computation, limited robustness in dynamic contexts, and lack of persistent autonomy and environmental awareness. We present Amico, a modular, event-driven framework for building autonomous agents optimized for embedded systems. Written in Rust for safety and performance, Amico supports reactive, persistent agents that operate efficiently across embedded platforms and browser environments via WebAssembly. It provides clean abstractions for event handling, state management, behavior execution, and integration with reasoning modules. Amico delivers a unified infrastructure for constructing resilient, interactive agents suitable for deployment in settings with limited compute and intermittent connectivity.

This paper presents the design, development, and application of a novel space simulation environment for rapidly prototyping and testing flight software for distributed space systems. The environment combines the flexibility, determinism, and observability of software-only simulation with the fidelity and depth normally attained only by real-time hardware-in-the-loop testing. Ultimately, this work enables an engineering process in which flight software is continuously improved and delivered in its final, flight-ready form, and which reduces the cost of design changes and software revisions with respect to a traditional linear development process. Three key methods not found in existing tools enable this environment's novel capabilities: first, a hybrid event-driven simulation architecture that combines continuous-time and discrete-event simulation paradigms; second, a lightweight application-layer software virtualization design that allows executing compiled flight software binaries while modeling process scheduling, input/output, and memory use; and third, high-fidelity models for the multi-spacecraft space environment, including for wireless communication, relative sensing such as differential GPS and cameras, and flight computer health metrics like heap exhaustion and fragmentation. The simulation environment's capabilities are applied to the iterative development and testing of two flight-ready software packages: the guidance, navigation, and control software for the VISORS mission, and the Stanford Space Rendezvous Laboratory software kit for rendezvous and proximity operations. Results from 33 months of flight software development demonstrate the use of this simulation environment to rapidly and reliably identify and resolve defects, characterize navigation and control performance, and scrutinize implementation details like memory allocation and inter-spacecraft network protocols.

Future multi-spacecraft missions require robust autonomous trajectory optimization capabilities to ensure safe and efficient rendezvous operations. This capability hinges on solving non-convex optimal control problems in real time, although traditional iterative methods such as sequential convex programming impose significant computational challenges. To mitigate this burden, the Autonomous Rendezvous Transformer introduced a generative model trained to provide near-optimal initial guesses. This approach provides convergence to better local optima (e.g., fuel optimality), improves feasibility rates, and results in faster convergence speed of optimization algorithms through warm-starting. This work extends the capabilities of ART to address robust chance-constrained optimal control problems. Specifically, ART is applied to challenging rendezvous scenarios in Low Earth Orbit (LEO), ensuring fault-tolerant behavior under uncertainty. Through extensive experimentation, the proposed warm-starting strategy is shown to consistently produce high-quality reference trajectories, achieving up to 30% cost improvement and 50% reduction in infeasible cases compared to conventional methods, demonstrating robust performance across multiple state representations. Additionally, a post hoc evaluation framework is proposed to assess the quality of generated trajectories and mitigate runtime failures, marking an initial step toward the reliable deployment of AI-driven solutions in safety-critical autonomous systems such as spacecraft.

Reliable and efficient trajectory optimization methods are a fundamental need for autonomous dynamical systems, effectively enabling applications including rocket landing, hypersonic reentry, spacecraft rendezvous, and docking. Within such safety-critical application areas, the complexity of the emerging trajectory optimization problems has motivated the application of AI-based techniques to enhance the performance of traditional approaches. However, current AI-based methods either attempt to fully replace traditional control algorithms, thus lacking constraint satisfaction guarantees and incurring in expensive simulation, or aim to solely imitate the behavior of traditional methods via supervised learning. To address these limitations, this paper proposes the Autonomous Rendezvous Transformer (ART) and assesses the capability of modern generative models to solve complex trajectory optimization problems, both from a forecasting and control standpoint. Specifically, this work assesses the capabilities of Transformers to (i) learn near-optimal policies from previously collected data, and (ii) warm-start a sequential optimizer for the solution of non-convex optimal control problems, thus guaranteeing hard constraint satisfaction. From a forecasting perspective, results highlight how ART outperforms other learning-based architectures at predicting known fuel-optimal trajectories. From a control perspective, empirical analyses show how policies learned through Transformers are able to generate near-optimal warm-starts, achieving trajectories that are (i) more fuel-efficient, (ii) obtained in fewer sequential optimizer iterations, and (iii) computed with an overall runtime comparable to benchmarks based on convex optimization.

This paper presents a neural network-based Unscented Kalman Filter (UKF) to estimate and track the pose (i.e., position and orientation) of a known, noncooperative, tumbling target spacecraft in a close-proximity rendezvous scenario. The UKF estimates the target's orbit and attitude relative to the servicer based on the pose information provided by a multi-task Convolutional Neural Network (CNN) from incoming monocular images of the target. In order to enable reliable tracking, the process noise covariance matrix of the UKF is tuned online using adaptive state noise compensation which leverages a newly developed closed-form process noise model for relative attitude dynamics. This paper also introduces the Satellite Hardware-In-the-loop Rendezvous Trajectories (SHIRT) dataset to enable comprehensive analyses of the performance and robustness of the proposed pipeline. SHIRT comprises the labeled images of two representative rendezvous trajectories in low Earth orbit created using both a graphics renderer and a robotic testbed. Specifically, the CNN is solely trained on synthetic data, whereas functionality and performance of the complete navigation pipeline are evaluated on real images from the robotic testbed. The proposed UKF is evaluated on SHIRT and is shown to have sub-decimeter-level position and degree-level orientation errors at steady-state.

Autonomous vision-based spaceborne navigation is an enabling technology for future on-orbit servicing and space logistics missions. While computer vision in general has benefited from Machine Learning (ML), training and validating spaceborne ML models are extremely challenging due to the impracticality of acquiring a large-scale labeled dataset of images of the intended target in the space environment. Existing datasets, such as Spacecraft PosE Estimation Dataset (SPEED), have so far mostly relied on synthetic images for both training and validation, which are easy to mass-produce but fail to resemble the visual features and illumination variability inherent to the target spaceborne images. In order to bridge the gap between the current practices and the intended applications in future space missions, this paper introduces SPEED+: the next generation spacecraft pose estimation dataset with specific emphasis on domain gap. In addition to 60,000 synthetic images for training, SPEED+ includes 9,531 hardware-in-the-loop images of a spacecraft mockup model captured from the Testbed for Rendezvous and Optical Navigation (TRON) facility. TRON is a first-of-a-kind robotic testbed capable of capturing an arbitrary number of target images with accurate and maximally diverse pose labels and high-fidelity spaceborne illumination conditions. SPEED+ is used in the second international Satellite Pose Estimation Challenge co-hosted by SLAB and the Advanced Concepts Team of the European Space Agency to evaluate and compare the robustness of spaceborne ML models trained on synthetic images.

Tiger Woods has told authorities he doesn't remember the rollover crash that landed him in a hospital with metal rods and pins in his leg. But the SUV he was driving does. Like other modern cars and trucks, the Genesis GV80 that Woods was driving when he crashed was equipped with an electronic data recorder and other computer hardware meant to serve as a digital witness of sorts -- filled with information investigators can use to piece together the seconds before and during the accident. The devices are part of a broader array of safety technology built into many newer vehicles. Vehicles in the Genesis line -- Hyundai's luxury brand -- for example, also feature artificial intelligence software that keeps a watchful eye, sending alerts if it detects the driver is distracted or closes his or her eyes while driving.

There is a large and growing environmental problem that is not on this Earth--it's trash orbiting our planet in space. Space debris is a not only an economic risk that can affect every day modern living, but also it is an existential risk that jeopardizes the ability for scientists to research weather and climate changes that impact all living things on Earth. In early February 2019, researchers from Stanford University's Space Rendezvous Laboratory (SLAB) announced a plan to crowdsource artificial intelligence (AI) to solve the massive problem of orbiting space debris. Simone D'Amico, founder and director of the Stanford's Space Rendezvous Lab, is partnering with the European Space Agency (ESA) to create an AI system that will provide navigational guidance to a space "tow truck" in order to identify, fix or remove defunct, orbiting satellites that are above the atmosphere, yet doomed to orbit due to the gravitational pull of the Earth. How can artificial intelligence solve the problem of space debris?

Intellivision, the video game maker that didn't survive the '80s, is back and ready to build something new on top of gaming nostalgia. After teasing the idea earlier this year, the company has announced plans for a new console called the Amico -- one part retro console and one part family-friendly modern gaming system. Intellivision plans to release it with a mix of classic titles and new originals in 2020. The Amico shares some similarities with the standard retro console release. It will supposedly play a slew of classic titles that '80s babies might remember playing through on their living room TV.