Search for: All records

Recent advances in computer vision for space exploration have handled prediction uncertainties well by approximating multimodal output distribution rather than averaging the distribution. While those advanced multimodal deep learning models could enhance the scientific and engineering value of autonomous systems by making the optimal decisions in uncertain environments, sequential learning of those approximated information has depended on unimodal or bimodal probability distribution. In a sequence of information learning and transfer decisions, the traditional reinforcement learning cannot accommodate the noise in the data that could be useful for gaining information from other locations, thus cannot handle multimodal and multivariate gains in their transition function. Still, there is a lack of interest in learning and transferring multimodal space information effectively to maximally remove the uncertainty. In this study, a new information theory overcomes the traditional entropy approach by actively sensing and learning information in a sequence. Particularly, the autonomous navigation of a team of heterogeneous unmanned ground and aerial vehicle systems in Mars outperforms benchmarks through indirect learning. 

Recent advances in computer vision for space exploration have handled prediction uncertainties well by approximating multimodal output distribution rather than averaging the distribution. While those advanced multimodal deep learning models could enhance the scientific and engineering value of autonomous systems by making the optimal decisions in uncertain environments, sequential learning of those approximated information has depended on unimodal or bimodal probability distribution. In a sequence of information learning and transfer decisions, the traditional reinforcement learning cannot accommodate the noise in the data that could be useful for gaining information from other locations, thus cannot handle multimodal and multivariate gains in their transition function. Still, there is a lack of interest in learning and transferring multimodal space information effectively to maximally remove the uncertainty. In this study, a new information theory overcomes the traditional entropy approach by actively sensing and learning information in a sequence. Particularly, the autonomous navigation of a team of heterogeneous unmanned ground and aerial vehicle systems in Mars outperforms benchmarks through indirect learning. 

We introduce temporal multimodal multivariate learning, a new family of decision making models that can indirectly learn and transfer online information from simultaneous observations of a probability distribution with more than one peak or more than one outcome variable from one time stage to another. We approximate the posterior by sequentially removing additional uncertainties across different variables and time, based on data-physics driven correlation, to address a broader class of challenging time-dependent decision-making problems under uncertainty. Extensive experiments on real-world datasets ( i.e., urban traffic data and hurricane ensemble forecasting data) demonstrate the superior performance of the proposed targeted decision-making over the state-of-the-art baseline prediction methods across various settings. 

Mission-critical exploration of uncertain environments requires reliable and robust mechanisms for achieving information gain. Typical measures of information gain such as Shannon entropy and KL divergence are unable to distinguish between different bimodal probability distributions or introduce bias toward one mode of a bimodal probability distribution. The use of a standard deviation (SD) metric reduces bias while retaining the ability to distinguish between higher and lower risk distributions. Areas of high SD can be safely explored through observation with an autonomous Mars Helicopter allowing safer and faster path plans for ground-based rovers. First, this study presents a single-agent information-theoretic utility-based path planning method for a highly correlated uncertain environment. Then, an information-theoretic two-stage multiagent rapidly exploring random tree framework is presented, which guides Mars helicopter through regions of high SD to reduce uncertainty for the rover. In a Monte Carlo simulation, we compare our information-theoretic framework with a rover-only approach and a naive approach, in which the helicopter scouts ahead of the rover along its planned path. Finally, the model is demonstrated in a case study on the Jezero region of Mars. Results show that the information-theoretic helicopter improves the travel time for the rover on average when compared with the rover alone or with the helicopter scouting ahead along the rover’s initially planned route. 

Lack of high‐resolution observations in the inner‐core of tropical cyclones remains a key issue when constructing an accurate initial state of the storm structure. The major implication of an improper initial state is the poor predictability of the future state of the storm. The size and associated hazard from strong winds at the inner‐core make it impossible to sample this region entirely. However, targeting regions of the inner‐core where forecasted atmospheric measurements have high uncertainty can significantly improve the accuracy of measurements for the initial state of the storm. This study provides a scheme for targeted high‐resolution observations for small Unmanned Aircraft Systems (sUAS) platforms (e.g., Coyote sUAS) to improve the estimates of the atmospheric measurement in the inner‐core structure. The benefit of observation is calculated based on the high‐fidelity state‐of‐the‐art hurricane ensemble data assimilation system. Potential locations with the mostinformativemeasurements are identified through exploration of various simulation‐based solutions depending on the state variables (e.g., pressure, temperature, wind speed, relative humidity) and a combined representation of those variables. A sampling‐based sUAS path planning algorithm considers energy usage when locating the regions of highly uncertain prediction of measurements, allowing sUAS to maximize the benefit of observation. Robustness analysis of our algorithm for multiple scenarios of sUAS drop and goal locations shows satisfactory performance against benchmark similar to current NOAA field campaign. With optimized sUAS observations, a data assimilation analysis shows significant improvements of up to 4% in the tropical cyclone structure estimates after resolving uncertainties at targeted locations.

 

Standard methods for synthesis of control policies in Markov decision processes with unknown transition probabilities largely rely on a combination of exploration and exploitation. While these methods often offer theoretical guarantees on system performance, the number of time steps and samples needed to initially explore the environment before synthesizing a well-performing control policy is impractically large. This paper partially alleviates such a burden by incorporating a priori existing knowledge into learning, when such knowledge is available. Based on prior information about bounds on the differences between the transition probabilities at different states, we propose a learning approach where the transition probabilities at a given state are not only learned from outcomes of repeatedly performing a certain action at that state, but also from outcomes of performing actions at states that are known to have similar transition probabilities. Since the directly obtained information is more reliable at determining transition probabilities than second-hand information, i.e., information obtained from similar but potentially slightly different states, samples obtained indirectly are weighted with respect to the known bounds on the differences of transition probabilities. While the proposed strategy can naturally lead to errors in learned transition probabilities, we show that, by proper choice of the weights, such errors can be reduced, and the number of steps needed to form a near-optimal control policy in the Bayesian sense can be significantly decreased. 

MAARS (Machine leaning-based Analytics for Automated Rover Systems) is an ongoing JPL effort to bring the latest self-driving technologies to Mars, Moon, and beyond. The ongoing AI revolution here on Earth is finally propagating to the red planet as the High Performance Spaceflight Computing (HPSC) and commercial off-the-shelf (COTS) system-on-a-chip (SoC), such as Qualcomm's Snapdragon, become available to rovers. In this three year project, we are developing, implementing, and benchmarking a wide range of autonomy algorithms that would significantly enhance the productivity and safety of planetary rover missions. This paper is to provide the latest snapshot of the project with broad and high-level description of every capability that we are developing, including scientific scene interpretation, vision-based traversability assessment, resource-aware path planning, information-theoretic path planning, on-board strategic path planning, and on-board optimal kinematic settling for accurate collision checking. All of the onboard software capabilities will be integrated into JPL's Athena test rover using ROS (Robot Operating System).