Publications

Refine Results

(Filters Applied) Clear All

Probabilistic coordination of heterogeneous teams from capability temporal logic specifications

Summary

This letter explores coordination of heterogeneous teams of agents from high-level specifications. We employ Capability Temporal Logic (CaTL) to express rich, temporal-spatial tasks that require cooperation between many agents with unique capabilities. CaTL specifies combinations of tasks, each with desired locations, duration, and set of capabilities, freeing the user from considering specific agent trajectories and their impact on multi-agent cooperation. CaTL also provides a quantitative robustness metric of satisfaction based on availability of required capabilities for each task. The novelty of this letter focuses on satisfaction of CaTL formulas under probabilistic conditions. Specifically, we consider uncertainties in robot motion (e.g., agents may fail to transition between regions with some probability) and local probabilistic workspace properties (e.g., if there are not enough agents of a required capability to complete a collaborative task). The proposed approach automatically formulates amixed-integer linear program given agents, their dynamics and capabilities, an abstraction of the workspace, and a CaTL formula. In addition to satisfying the given CaTL formula, the optimization considers the following secondary goals (in decreasing order of priority): 1) minimize the risk of transition failure due to uncertainties; 2) maximize probabilities of regional collaborative satisfaction (if there is an excess of agents); 3) maximize the availability robustness of CaTL for potential agent attrition; 4) minimize the total agent travel time. We evaluate the performance of the proposed framework and demonstrate its scalability via numerical simulations.
READ LESS

Summary

This letter explores coordination of heterogeneous teams of agents from high-level specifications. We employ Capability Temporal Logic (CaTL) to express rich, temporal-spatial tasks that require cooperation between many agents with unique capabilities. CaTL specifies combinations of tasks, each with desired locations, duration, and set of capabilities, freeing the user from...

READ MORE

Fast decomposition of temporal logic specifications for heterogeneous teams

Published in:
IEEE Robot. Autom. Lett., Vol. 7, No. 2, April 2022, pp. 2297-2304.

Summary

We focus on decomposing large multi-agent path planning problems with global temporal logic goals (common to all agents) into smaller sub-problems that can be solved and executed independently. Crucially, the sub-problems' solutions must jointly satisfy the common global mission specification. The agents' missions are given as Capability Temporal Logic (CaTL) formulas, a fragment of Signal Temporal Logic (STL) that can express properties over tasks involving multiple agent capabilities (i.e., different combinations of sensors, effectors, and dynamics) under strict timing constraints. We jointly decompose both the temporal logic specification and the team of agents, using a satisfiability modulo theories (SMT) approach and heuristics for handling temporal operators. The output of the SMT is then distributed to subteams and leads to a significant speed up in planning time compared to planning for the entire team and specification. We include computational results to evaluate the efficiency of our solution, as well as the trade-offs introduced by the conservative nature of the SMT encoding and heuristics.
READ LESS

Summary

We focus on decomposing large multi-agent path planning problems with global temporal logic goals (common to all agents) into smaller sub-problems that can be solved and executed independently. Crucially, the sub-problems' solutions must jointly satisfy the common global mission specification. The agents' missions are given as Capability Temporal Logic (CaTL)...

READ MORE

Quantifying bias in face verification system

Summary

Machine learning models perform face verification (FV) for a variety of highly consequential applications, such as biometric authentication, face identification, and surveillance. Many state-of-the-art FV systems suffer from unequal performance across demographic groups, which is commonly overlooked by evaluation measures that do not assess population-specific performance. Deployed systems with bias may result in serious harm against individuals or groups who experience underperformance. We explore several fairness definitions and metrics, attempting to quantify bias in Google’s FaceNet model. In addition to statistical fairness metrics, we analyze clustered face embeddings produced by the FV model. We link well-clustered embeddings (well-defined, dense clusters) for a demographic group to biased model performance against that group. We present the intuition that FV systems underperform on protected demographic groups because they are less sensitive to differences between features within those groups, as evidenced by clustered embeddings. We show how this performance discrepancy results from a combination of representation and aggregation bias.
READ LESS

Summary

Machine learning models perform face verification (FV) for a variety of highly consequential applications, such as biometric authentication, face identification, and surveillance. Many state-of-the-art FV systems suffer from unequal performance across demographic groups, which is commonly overlooked by evaluation measures that do not assess population-specific performance. Deployed systems with bias...

READ MORE

Correlated Bayesian model of aircraft encounters in the terminal area given a straight takeoff or landing

Published in:
Aerospace, Vol. 9, No.2, 12 March 2022.

Summary

The integration of new airspace entrants into terminal operations requires design and evaluation of Detect and Avoid systems that prevent loss of well clear from and collision with other aircraft. Prior to standardization or deployment, an analysis of the safety performance of those systems is required. This type of analysis has typically been conducted by Monte Carlo simulation with synthetic, statistically representative encounters between aircraft drawn from an appropriate encounter model. While existing encounter models include terminal airspace classes, none explicitly represents the structure expected while engaged in terminal operations, e.g., aircraft in a traffic pattern. The work described herein is an initial model of such operations where an aircraft landing or taking off via a straight trajectory encounters another aircraft landing or taking off, or transiting by any means. The model shares the Bayesian network foundation of other Massachusetts Institute of Technology Lincoln Laboratory encounter models but tailors those networks to address structured terminal operations, i.e., correlations between trajectories and the airfield and each other. This initial model release is intended to elicit feedback from the standards-writing community.
READ LESS

Summary

The integration of new airspace entrants into terminal operations requires design and evaluation of Detect and Avoid systems that prevent loss of well clear from and collision with other aircraft. Prior to standardization or deployment, an analysis of the safety performance of those systems is required. This type of analysis...

READ MORE

Robust network protocols for large swarms of small UAVs

Summary

In this work, we detail a synchronized channel hopping network for autonomous swarms of small unmanned aerial vehicles (UAVs) conducting intelligence, surveillance, and reconnaissance (ISR) missions in the presence of interference and jamming. The core component of our design is Queue Length Informed Maximal Matching (QLIMM), a distributed transmission scheduling protocol that exchanges queue state information between nodes to assign subdivisions of the swarm to orthogonal hopping patterns in response to the network’s throughput demands. QLIMM efficiently allocates channel resources across large networks without relying on any centralized control or pre-planned traffic patterns, which is in the spirit of a swarming capability. However, given that the control messaging must scale up with the swarm’s size and the challenging interference environments we consider, fragility could be a concern. To observe under what conditions control fails, we test our protocol against both simulated partial-band noise jamming and background interference. For the latter, we use data collected from a small unmanned aircraft system to characterize the interference seen by a UAV in the 2.4 and 5 GHz bands in both urban and rural settings. These measurements show that the interference can be 15 dB higher at a 50-meter flight altitude when compared to observations on the ground. Using this data, we conduct extensive network simulations of QLIMM in Riverbed Modeler to show that, under moderate jamming and interference, it outperforms traditional channel access methods as well as other scheduling protocols that do not pass queue state information.
READ LESS

Summary

In this work, we detail a synchronized channel hopping network for autonomous swarms of small unmanned aerial vehicles (UAVs) conducting intelligence, surveillance, and reconnaissance (ISR) missions in the presence of interference and jamming. The core component of our design is Queue Length Informed Maximal Matching (QLIMM), a distributed transmission scheduling...

READ MORE

Radar coverage analysis for the Terminal Precipitation on the Glass Program

Author:
Published in:
MIT Lincoln Laboratory Report ATC-450

Summary

The Terminal Precipitation on the Glass (TPoG) program proposes to improve the STARS precipitation depiction by adding an alternative precipitation product based on a national weather-radar-based mosaic, i.e., the NextGen Weather System (aka NextGen Weather Processor [NWP] and Common Support Services Weather [CSS-Wx]). This report describes spatial and temporal domain analyses conducted over the 146 terminal radar approach control (TRACON) airspaces that are within scope of TPoG to identify and quantify future TPoG benefits, as well as potential operational issues.
READ LESS

Summary

The Terminal Precipitation on the Glass (TPoG) program proposes to improve the STARS precipitation depiction by adding an alternative precipitation product based on a national weather-radar-based mosaic, i.e., the NextGen Weather System (aka NextGen Weather Processor [NWP] and Common Support Services Weather [CSS-Wx]). This report describes spatial and temporal domain...

READ MORE

Detection of COVID-19 using multimodal data from a wearable device: results from the first TemPredict Study

Summary

Early detection of diseases such as COVID-19 could be a critical tool in reducing disease transmission by helping individuals recognize when they should self-isolate, seek testing, and obtain early medical intervention. Consumer wearable devices that continuously measure physiological metrics hold promise as tools for early illness detection. We gathered daily questionnaire data and physiological data using a consumer wearable (Oura Ring) from 63,153 participants, of whom 704 self-reported possible COVID-19 disease. We selected 73 of these 704 participants with reliable confirmation of COVID-19 by PCR testing and high-quality physiological data for algorithm training to identify onset of COVID-19 using machine learning classification. The algorithm identified COVID-19 an average of 2.75 days before participants sought diagnostic testing with a sensitivity of 82% and specificity of 63%. The receiving operating characteristic (ROC) area under the curve (AUC) was 0.819 (95% CI [0.809, 0.830]). Including continuous temperature yielded an AUC 4.9% higher than without this feature. For further validation, we obtained SARS CoV-2 antibody in a subset of participants and identified 10 additional participants who self-reported COVID-19 disease with antibody confirmation. The algorithm had an overall ROC AUC of 0.819 (95% CI [0.809, 0.830]), with a sensitivity of 90% and specificity of 80% in these additional participants. Finally, we observed substantial variation in accuracy based on age and biological sex. Findings highlight the importance of including temperature assessment, using continuous physiological features for alignment, and including diverse populations in algorithm development to optimize accuracy in COVID-19 detection from wearables.
READ LESS

Summary

Early detection of diseases such as COVID-19 could be a critical tool in reducing disease transmission by helping individuals recognize when they should self-isolate, seek testing, and obtain early medical intervention. Consumer wearable devices that continuously measure physiological metrics hold promise as tools for early illness detection. We gathered daily...

READ MORE

Bayesian estimation of PLDA in the presence of noisy training labels, with applications to speaker verification

Published in:
IEEE/ACM Trans. Audio, Speech, Language Process., Vol. 30, 2022, pp. 414-28.

Summary

This paper presents a Bayesian framework for estimating a Probabilistic Linear Discriminant Analysis (PLDA) model in the presence of noisy labels. True class labels are interpreted as latent random variables, which are transmitted through a noisy channel, and received as observed speaker labels. The labeling process is modeled as a Discrete Memoryless Channel (DMC). PLDA hyperparameters are interpreted as random variables, and their joint posterior distribution is derived using meanfield Variational Bayes, allowing maximum a posteriori (MAP) estimates of the PLDA model parameters to be determined. The proposed solution, referred to as VB-MAP, is presented as a general framework, but is studied in the context of speaker verification, and a variety of use cases are discussed. Specifically, VB-MAP can be used for PLDA estimation with unreliable labels, unsupervised PLDA estimation, and to infer the reliability of a PLDA training set. Experimental results show the proposed approach to provide significant performance improvements on a variety of NIST Speaker Recognition Evaluation (SRE) tasks, both for data sets with simulated mislabels, and for data sets with naturally occurring missing or unreliable labels.
READ LESS

Summary

This paper presents a Bayesian framework for estimating a Probabilistic Linear Discriminant Analysis (PLDA) model in the presence of noisy labels. True class labels are interpreted as latent random variables, which are transmitted through a noisy channel, and received as observed speaker labels. The labeling process is modeled as a...

READ MORE

Energy resilience: exercises for Marine Corps installations

Published in:
Marine Corps Gazette, Vol. 106, No. 2, February 2022, p. 20-24.
Topic:
R&D group:

Summary

Microgrids are areas that are self-sufficient for power that can controllably disconnect from the incoming utility feed and control generation assets in conjunction with changing load requirements. They are increasingly being touted as a way to improve installations energy resilience because they allow installations to decouple from the larger electric grid if it fails and continue to provide power in the face of growing natural and man-made threats to Marine Corps installations. However, before commanders can put resources toward upgrading infrastructure, they need to identify and understand their vulnerabilities. A key way to do this is by holding exercises designed to simulate grid failures and outages either in a tabletop manner or in realtime. These exercises also help personnel train for disruptions, understand their impact on operations, and identify unknown interdependencies that can be just as important as investing in resilient technology and the physical electric grid. In order for the equipment to work, personnel have to know how to employ it and commands need to understand how outages will affect their installations. These types of exercises are as important as the physical infrastructure or ensuring the energy resilience of Marine Corps installations and the missions that depend on them in the future.
READ LESS

Summary

Microgrids are areas that are self-sufficient for power that can controllably disconnect from the incoming utility feed and control generation assets in conjunction with changing load requirements. They are increasingly being touted as a way to improve installations energy resilience because they allow installations to decouple from the larger electric...

READ MORE

Tools and practices for responsible AI engineering

Summary

Responsible Artificial Intelligence (AI)—the practice of developing, evaluating, and maintaining accurate AI systems that also exhibit essential properties such as robustness and explainability—represents a multifaceted challenge that often stretches standard machine learning tooling, frameworks, and testing methods beyond their limits. In this paper, we present two new software libraries—hydra-zen and the rAI-toolbox—that address critical needs for responsible AI engineering. hydra-zen dramatically simplifies the process of making complex AI applications configurable, and their behaviors reproducible. The rAI-toolbox is designed to enable methods for evaluating and enhancing the robustness of AI-models in a way that is scalable and that composes naturally with other popular ML frameworks. We describe the design principles and methodologies that make these tools effective, including the use of property-based testing to bolster the reliability of the tools themselves. Finally, we demonstrate the composability and flexibility of the tools by showing how various use cases from adversarial robustness and explainable AI can be concisely implemented with familiar APIs.
READ LESS

Summary

Responsible Artificial Intelligence (AI)—the practice of developing, evaluating, and maintaining accurate AI systems that also exhibit essential properties such as robustness and explainability—represents a multifaceted challenge that often stretches standard machine learning tooling, frameworks, and testing methods beyond their limits. In this paper, we present two new software libraries—hydra-zen and...

READ MORE