Emergent Dynamics, Control and Analytics Labs

Filter by type:

Sort by year:

Sensor-Based Calibration of Camera's Extrinsic Parameters for Stereophotogrammetry

Fabio Bottalico, Christopher Niezrecki, Kshitij Jerath, Yan Luo, and Alessandro Sabato.
Journal PaperIEEE Sensors Journal, DOI: 10.1109/JSEN.2023.3244413 (2023)

Abstract

Stereophotogrammetry is a well-recognized technique for structural health monitoring. Before performing any quantitative stereophotogrammetry measurement, the cameras must be calibrated to obtain the intrinsic and extrinsic parameters of the stereovision system. However, when large-sized structures are to be monitored, the calibration process is challenging and limits the use of stereophotogrammetry. In this research, a sensor-based calibration method for determining the extrinsic parameters of a stereovision system is presented and validated. A multi-sensor board has been developed that synchronizes inertial measurement units (IMUs) and a laser on a single board computer to measure the spatial orientation and the distance of two paired cameras and compute the extrinsic parameters of the stereovision system. The effectiveness of the sensor-based calibration is evaluated through both analytical studies to quantify the effects of performance degradation caused by the sensors' noise as well as laboratory tests. Results show that the sensor-based calibration is effective in quantifying displacement with errors below 3% when compared to measurements performed using a stereovision system calibrated with the traditional image-based procedure.

Granulation of Large Temporal Databases: An Allan Variance Approach

Lorina Sinanaj#, Hossein Haeri*, Satya Maddipatla, Liming Gao, Rinith Pakala#, Niket Kathiriya*, Craig Beal, Sean Brennan, Cindy Chen, and Kshitij Jerath
Journal PaperSpringer Nature Computer Science vol. 4, no. 7 (2023)

Abstract

As the use of Big Data begins to dominate various scientific and engineering applications, the ability to conduct complex data analyses with speed and efficiency has become increasingly important. The availability of large amounts of data results in ever-growing storage requirements and magnifies issues related to query response times. In this work, we propose a novel methodology for granulation and data reduction of large temporal databases that can address both issues simultaneously. While prior data reduction techniques rely on heuristics or may be computationally intensive, our work borrows the concept of Allan Variance (AVAR) from the fields of signal processing and sensor characterization to efficiently and systematically reduce the size of temporal databases. Specifically, we use Allan variance to systematically determine the temporal window length over which data remains relevant. Large temporal databases are then granulated using the AVAR-determined window length. Averaging over the resulting granules produces aggregate information for each granule, resulting in significant data reduction. The query performance and data quality are evaluated using existing standard datasets, as well as for two large datasets that include temporal information for vehicular and weather data. Our results demonstrate that the AVAR-based data reduction approach is efficient and maintains data quality, while leading to an order of magnitude improvement in query execution times compared to three existing clustering-based data reduction methods.

DECISIVE Benchmarking Data Report: sUAS Performance Results from Phase I

Adam Norton, Reza Ahmadzadeh, Kshitij Jerath, Paul Robinette, Jay Weitzen, Thanuka Wickramarathne, Holly Yanco, Minseop Choi, Ryan Donald, Brendan Donoghue, Christian Dumas, Peter Gavriel, Alden Giedraitis, Brendan Hertel, Jack Houle, Nathan Letteri, Edwin Meriaux, Zahra Rezaei Khavas, Rakshith Singh, Gregg Willcox, and Naye Yoni
Technical reportU.S. Army Combat Capabilities Development Command Soldier Center (DEVCOM-SC); Contract # W911QY-18-2-0006

Abstract

This handbook outlines all test methods developed under the Development and Execution of Comprehensive and Integrated Subterranean Intelligent Vehicle Evaluations (DECISIVE) project by the University of Massachusetts Lowell for evaluating small unmanned aerial systems (sUAS) performance in subterranean and constrained indoor environments, spanning communications, field readiness, interface, obstacle avoidance, navigation, mapping, autonomy, trust, and situation awareness. For sUAS deployment in subterranean and constrained indoor environments, this puts forth two assumptions about applicable sUAS to be evaluated using these test methods: (1) able to operate without access to GPS signal, and (2) width from prop top to prop tip does not exceed 91 cm (36 in) wide (i.e., can physically fit through a typical doorway, although successful navigation through is not guaranteed). All test methods are specified using a common format: Purpose, Summary of Test Method, Apparatus and Artifacts, Equipment, Metrics, Procedure, and Example Data. All test methods are designed to be run in real-world environments (e.g., MOUT sites) or using fabricated apparatuses (e.g., test bays built from wood, or contained inside of one or more shipping containers).

DECISIVE Test Methods Handbook: Test Methods for Evaluating sUAS in Subterranean and Constrained Indoor Environments, Version 1.1

Adam Norton, Reza Ahmadzadeh, Kshitij Jerath, Paul Robinette, Jay Weitzen, Thanuka Wickramarathne, Holly Yanco, Minseop Choi, Ryan Donald, Brendan Donoghue, Christian Dumas, Peter Gavriel, Alden Giedraitis, Brendan Hertel, Jack Houle, Nathan Letteri, Edwin Meriaux, Zahra Rezaei Khavas, Rakshith Singh, Gregg Willcox, and Naye Yoni
Techincal reportU.S. Army Combat Capabilities Development Command Soldier Center (DEVCOM-SC); Contract # W911QY-18-2-0006

Abstract

This handbook outlines all test methods developed under the Development and Execution of Comprehensive and Integrated Subterranean Intelligent Vehicle Evaluations (DECISIVE) project by the University of Massachusetts Lowell for evaluating small unmanned aerial systems (sUAS) performance in subterranean and constrained indoor environments, spanning communications, field readiness, interface, obstacle avoidance, navigation, mapping, autonomy, trust, and situation awareness. For sUAS deployment in subterranean and constrained indoor environments, this puts forth two assumptions about applicable sUAS to be evaluated using these test methods: (1) able to operate without access to GPS signal, and (2) width from prop top to prop tip does not exceed 91 cm (36 in) wide (i.e., can physically fit through a typical doorway, although successful navigation through is not guaranteed). All test methods are specified using a common format: Purpose, Summary of Test Method, Apparatus and Artifacts, Equipment, Metrics, Procedure, and Example Data. All test methods are designed to be run in real-world environments (e.g., MOUT sites) or using fabricated apparatuses (e.g., test bays built from wood, or contained inside of one or more shipping containers).

Understanding Agent Competency: Effects of Environment Complexity on Area Coverage Time

Fernando Mazzoni* and Kshitij Jerath
Conference PaperIntelligent Agents (IA) at IEEE Symposium Series on Computational Intelligence 2022, Singapore, Dec 4-7, 2022

Abstract

As an increasing number of search-and-rescue (SAR) missions begin to incorporate robotic agents and algorithms to cover search areas, the task of selecting an appropriate algorithm or robot for a specific mission becomes critical. This is particularly challenging since the search time or area coverage performances of SAR algorithms may vary significantly from one operating environment to another. While previous works have evaluated the performance of various SAR algorithms, few have examined how this performance changes as a function of the complexity of the search area or environment. In the presented work, we seek to overcome this knowledge gap by (a) developing the notion of algorithm competency, and (b) generating a competency profile for a specific biologically-inspired SAR algorithm (Lévy walk search) that relates the algorithm's performance to the complexity of the search environment. Our simulation-based study leverages several existing measures of network topological complexity. Our results show that we can successfully generate competency profiles that indicate positive correlation between the area coverage time and certain complexity measures of the environment. We have also found that some algorithms work better in low complexity environments, while other algorithms outperform them in more complex environments, indicating that competency profiles can aid in the selection of appropriate algorithms.

Evaluation of Navigation and Trajectory-following Capabilities of Small Unmanned Aerial System

Edwin Meriaux* and Kshitij Jerath
Conference Paper2022 IEEE Symposium on Technologies for Homeland Security

Abstract

Use cases for Small Unmanned Aerial Systems (sUAS) have expanded significantly over the past few years. One use case that is relevant to both civilian and defense missions is reliable operation in GPS-denied indoor and subterranean (subT) environments such as urban underground, tunnel systems, and cave networks. While many sUAS evaluation studies exist for outdoor environments, there have been limited studies to evaluate the characteristics of sUAS in GPS-denied indoor and subT environments. This paper attempts to resolve this knowledge gap by presenting a methodology for evaluating the navigation performance of sUAS in such environments, including operations such as waypoint navigation, path traversal, trajectory keeping, and navigation around corners. Specifically, we determine and present results for the navigation performance of five commercially available sUAS via the presented evaluation methodology.

A sensor-based calibration system for three-dimensional digital image correlation

Fabio Bottalico, Nicholas A. Valente, Shweta Dabetwar, Kshitij Jerath, Yan Luo, Christopher Niezrecki, and Alessandro Sabato
Conference PaperProc. SPIE 12048, Health Monitoring of Structural and Biological Systems XVI, 120480Z (19 April 2022)

Abstract

Three-dimensional digital image correlation (3D-DIC) has become a strong alternative to traditional contact-based techniques for structural health monitoring. 3D-DIC can extract the full-field displacement of a structure from a set of synchronized stereo images. Before performing 3D-DIC, a complex calibration process must be completed to obtain the stereovision system’s extrinsic parameters (i.e., cameras’ distance and orientation). The time required for the calibration depends on the dimensions of the targeted structure. For example, for large-scale structures, the calibration may take several hours. Furthermore, every time the cameras’ position changes, a new calibration is required to recalculate the extrinsic parameters. The approach proposed in this research allows determining the 3D-DIC extrinsic parameters using the data measured with commercially available sensors. The system utilizes three Inertial Measurement Units with a laser distance meter to compute the relative orientation and distance between the cameras. In this paper, an evaluation of the sensitivity of the newly developed sensor suite is provided by assessing the errors in the measurement of the extrinsic parameters. Analytical simulations performed on a 7.5 x 5.7 m field of view using the data retrieved from the sensors show that the proposed approach provides an accuracy of ~10-6 m and a promising way to reduce the complexity of 3D-DIC calibration.

Reward-Sharing Relational Networks in Multi-Agent Reinforcement Learning as a Framework for Emergent Behavior

Hossein Haeri, Reza Ahmadzadeh, and Kshitij Jerath
PosterNortheastern Robotics Colloquium (NERC) 2022, Lowell, MA, 2022

Understanding Agent Competency: Effects of Environment Complexity on Area Coverage Time

Alden Daniels, Fernando Mazzoni, and Kshitij Jerath
PosterNortheastern Robotics Colloquium (NERC) 2022, Lowell, MA, 2022

Optimal Moving Average Estimation of Noisy Random Walks using Allan Variance-informed Window Length

Hossein Haeri*, Behrad Soleimani, and Kshitij Jerath
Conference Paper2022 American Control Conference (ACC), Atlanta, GA, USA, 2022, pp. 1646-1651

Abstract

Moving averages are widely used to estimate time-varying parameters, especially when the underlying dynamic model is unknown or uncertain. However, the selection of the optimal window length over which to evaluate the moving averages remains an unresolved issue in the field. In this paper, we demonstrate the use of Allan variance to identify the characteristic timescales of a noisy random walk from historical measurements. Further, we provide a closed-form, analytical result to show that the Allan variance-informed averaging window length is indeed the optimal averaging window length in the context of moving average estimation of noisy random walks. We complement the analytical proof with numerical results that support the solution, which is also reflected in the authors’ related works. This systematic methodology for selecting the optimal averaging window length using Allan variance is expected to widely benefit practitioners in a diverse array of fields that utilize the moving average estimation technique for noisy random walk signals.

Boxes-based Representation and Data Sharing of Road Surface Friction for CAVs

Liming Gao, Craig Beal, Wushuang Bai, Satya Prasad Maddipatla, Cindy Chen, Kshitij Jerath, Hossein Haeri*, Lorina Sinanaj*, and Sean Brennan
Conference Paper2022 Road Safety and Simulation International Conference (RSS)

Abstract

This work develops algorithms demonstrating fast implementations of Allan variance (AVAR) for regularly and irregularly sampled signals. AVAR is a technique first developed to study the frequency stability of atomic clocks. Typical AVAR algorithms calculate changes in means between differently-sized groupings of data and thus are useful in many data aggregation processes: to select the appropriate window length or timescales for estimating a signal’s moving average, to find the minimum variance of a signal, or to estimate the change in variance of a signal with complex noise contributions as a function of the number of collected data points. Unfortunately, AVAR typically involves very large signal lengths, yet the typical time required to compute AVAR increases quickly with the length of the time-series data. This paper presents a recursive algorithm inspired by the Fast Fourier Transform (FFT), specifically data organization into power-of-two groupings. This enables a fast AVAR implementation, called FAVAR, shown first for regularly sampled data. The results show a computational speed increase of three orders of magnitude versus typical AVAR calculations for data lengths often used with AVAR. Next, the FAVAR algorithm is extended to compute AVAR of irregularly sampled data by modeling these data as weighted but regularly sampled data clusters. Finally, this work analyzes Dynamic Allan variance implementations of FAVAR, called D-FAVAR, wherein AVAR is calculated at every timestep to capture window-varying statistical properties of the data stream. The recursion methods used in FAVAR, when extended to compute D-FAVAR, further increase computational speed by an additional factor of ten compared to computing the FAVAR at every timestep. They result in approximately four orders of magnitude speed improvements versus repeated calculation of AVAR with typical methods. These fast algorithms are demonstrated on signals that illustrate classical Allan variance curves, and the results agree with the classical AVAR formulations within computational accuracy.

Macroscopic Observability of Emergent Behaviors in Multi-scale Models of Traffic Flow

Zhaohui Yang*
Dissertation Ph.D., Mechanical Engineering, University of Massachusetts Lowell, Feb 2022

Abstract

Emergent behaviors in complex systems arise due to the numerous local interactions that occur between large numbers of individual agents. The observation and quantification of such emergent behaviors remains a challenging task as information from all agents in the system is not always readily available. These challenges are also evident in the context of large-scale traffic systems, where the task of using measurements from a few individual vehicles to effectively observe and predict emergent congestion patterns (such as backward moving waves, and self-organizing or `phantom' traffic jams) remains currently unresolved. Attempts to understand the macroscopic observability of emergent traffic flows could benefit from a single framework that can seamlessly model traffic across different spatial scales, enabling the use of microscopic measurements made by vehicles to evaluate observability at another modeling scale. The presented work demonstrates how trends in observability of emergent behaviors can be determined as a function of the order of the system model. Specifically, it is shown that a trade-off exists between accuracy of the model and its observability, which may enable us to choose a `favorable' modeling scale that balances the advantages offered by increased observability versus increased accuracy. However, the control-theoretic model order reduction technique was found to be difficult to scale to larger traffic systems, so an alternative statistical mechanics-based approach using renormalization group (RG) theory is also presented and examined. The study indicates that this approach enables us to generate models of traffic flow dynamics at several coarser spatial scales in a manner such that the parameters of the models at different scales can be analytically related. The Renormalization Group-theoretic approach was further built upon to create a mutual information-based macroscopic observability metric that quantified the ability to relate the measurements made by an individual agent at the microscopic scale to the likelihood of occurrence of specific macroscopic states (i.e. emergent behaviors) of the traffic flow system. This approach could enable researchers and practitioners in transportation engineering and complex systems science to better understand macroscopic scale dynamics using microscopic scale information, potentially spurring advances in prediction and control of emergent behaviors.

Fast Allan Variance (FAVAR) and Dynamic Fast Allan Variance (D-FAVAR) Algorithms for both Regularly and Irregularly Sampled Data

Satya Prasad Maddipatla, Hossein Haeri*, Kshitij Jerath, and Sean Brennan
Journal Paper IFAC-PapersOnLine 54, no. 20 (2021): 26-31.

Abstract

This work develops algorithms demonstrating fast implementations of Allan variance (AVAR) for regularly and irregularly sampled signals. AVAR is a technique first developed to study the frequency stability of atomic clocks. Typical AVAR algorithms calculate changes in means between differently-sized groupings of data and thus are useful in many data aggregation processes: to select the appropriate window length or timescales for estimating a signal’s moving average, to find the minimum variance of a signal, or to estimate the change in variance of a signal with complex noise contributions as a function of the number of collected data points. Unfortunately, AVAR typically involves very large signal lengths, yet the typical time required to compute AVAR increases quickly with the length of the time-series data. This paper presents a recursive algorithm inspired by the Fast Fourier Transform (FFT), specifically data organization into power-of-two groupings. This enables a fast AVAR implementation, called FAVAR, shown first for regularly sampled data. The results show a computational speed increase of three orders of magnitude versus typical AVAR calculations for data lengths often used with AVAR. Next, the FAVAR algorithm is extended to compute AVAR of irregularly sampled data by modeling these data as weighted but regularly sampled data clusters. Finally, this work analyzes Dynamic Allan variance implementations of FAVAR, called D-FAVAR, wherein AVAR is calculated at every timestep to capture window-varying statistical properties of the data stream. The recursion methods used in FAVAR, when extended to compute D-FAVAR, further increase computational speed by an additional factor of ten compared to computing the FAVAR at every timestep. They result in approximately four orders of magnitude speed improvements versus repeated calculation of AVAR with typical methods. These fast algorithms are demonstrated on signals that illustrate classical Allan variance curves, and the results agree with the classical AVAR formulations within computational accuracy.

Allan Variance-based Granulation Technique for Large Temporal Databases

Lorina Sinanaj*, Hossein Haeri*, Liming Gao, Satya Maddipatla, Cindy Chen, Kshitij Jerath, Craig Beal, and Sean Brennan.
Conference Paper 13th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management-KMIS. 2021.

Abstract

In the era of Big Data, conducting complex data analysis tasks efficiently, becomes increasingly important and challenging due to large amounts of data available. In order to decrease query response time with limited main memory and storage space, data reduction techniques that preserve data quality are needed. Existing data reduction techniques, however, are often computationally expensive and rely on heuristics for deciding how to split or reduce the original dataset. In this paper, we propose an effective granular data reduction technique for temporal databases, based on Allan Variance (AVAR). AVAR is used to systematically determine the temporal window length over which data remains relevant. The entire dataset to be reduced is then separated into granules with size equal to the AVAR-determined window length. Data reduction is achieved by generating aggregated information for each such granule. The proposed method is tested using a large database that contains temporal information for vehicular data. Then comparison experiments are conducted and the outstanding runtime performance is illustrated by comparing with three clustering-based data reduction methods. The performance results demonstrate that the proposed Allan Variance-based technique can efficiently generate reduced representation of the original data without losing data quality, while significantly reducing computation time.

Reward-Sharing Relational Networks in Multi-Agent Reinforcement Learning as a Framework for Emergent Behavior

Hossein Haeri*, Reza Ahmadzadeh, and Kshitij Jerath
Workshop Paper Adaptive and Learning Agents Workshop at AAMAS 2021

Abstract

In this work, we integrate 'social' interactions into the multi-agent reinforcement learning (MARL) setup through a user-defined relational network and examine the effects of agent-agent relations on the rise of emergent behaviors. Leveraging insights from sociology and neuroscience, our proposed framework models agent relationships using the notion of Reward-Sharing Relational Networks (RSRN), where network edge weights act as a measure of how much one agent is invested in the success of (or `cares about') another. We construct relational rewards as a function of the RSRN interaction weights to collectively train the multi-agent system via a multi-agent reinforcement learning algorithm. The performance of the system is tested for a 3-agent scenario with different relational network structures (e.g., self-interested, communitarian, and authoritarian networks). Our results indicate that reward-sharing relational networks can significantly influence learned behaviors. We posit that RSRN can act as a framework where different relational networks produce distinct emergent behaviors, often analogous to the intuited sociological understanding of such networks.

A Micro-simulation Framework for Studying CAVs Behavior and Control Utilizing a Traffic Simulator, Chassis Simulation, and a Shared Roadway Friction Database

Liming Gao, Satya Maddipatla, Craig Beal, Kshitij Jerath, Cindy Chen, Lorina Sinanaj*, Hossein Haeri*, and Sean Brennan
Conference Paper 2021 American Control Conference (ACC), New Orleans, LA, USA, 2021, pp. 1650-1655.

Abstract

The ability of connected and autonomous vehicles (CAVs) to share information such as road friction and geometry has the potential to improve the safety, capacity, and efficiency of roadway systems, and the study of these systems often necessitates synergistic investigation of the vehicle, traffic behavior, and road conditions. This paper presents a micro-simulation framework for studying CAVs behavior and control utilizing a traffic simulator, chassis simulation, and a shared roadway friction database. The simulation utilizes three levels of data representations: 1) a traffic representation that explains how vehicles interact with each other and follow location-specific rules of the road, 2) a vehicle dynamic representation of the Newtonian response of the vehicle to driver inputs interacting with the vehicle which in turn interacts with the pavement, and finally 3) a road surface representation that represents how friction of roadway changes with space and time. The interactions between these representations are mediated by a spatiotemporal database. The framework is demonstrated through a CAVs application example showing how the mapping of road friction enables advanced vehicle control by allowing the database-mediated preview of road friction. This framework extends readily to real-time implementation on actual CAVs systems, providing great potential for improving CAVs control performance and stability via database-mediated feedback systems, not only in simulation, but also in practice.

Congestion-Aware Cooperative Adaptive Cruise Control for Mitigation of Self-Organized Traffic Jams

Taehooie Kim and Kshitij Jerath
Journal PaperIEEE Transactions on Intelligent Transportation Systems, vol. 23, no. 7, pp. 6621-6632, July 2022

Abstract

Previous work has shown that Adaptive Cruise Control (ACC) can improve traffic flow by raising the critical vehicular density at which congestion first appears. However, these works also indicate that traffic with medium-to-high penetration of ACC-equipped vehicles is more susceptible to the formation of self-organized phantom traffic jams induced by perturbations in vehicular demographics. In this work, we propose a congestion-aware Cooperative Adaptive Cruise Control (CACC) algorithm as an alternative to address the trade-off between competing goals of raising the critical density, and reducing susceptibility to congestion observed at higher penetration rates of ACC-equipped vehicles. The congestion-aware CACC algorithm is modeled after the General Motor’s car-following models, wherein the driver sensitivity is altered based on the prevailing congestion state (or traffic jam size) downstream of the connected vehicle. The dynamics of the self-organized traffic jam are modeled using a master equation. Results indicate that the congestion-aware CACC algorithm can increase the effective critical density leading to higher traffic flows, while also reducing the susceptibility to perturbations in vehicle demographics in the density range that adversely affects ACC-equipped vehicles.

Renormalization Group Approach to Cellular Automata-based Multi-scale Modeling of Traffic Flow

Zhaohui Yang* and Kshitij Jerath
Conference Paper Unifying Themes in Complex Systems X: Proceedings of the Tenth International Conference on Complex Systems (Springer Proceedings in Complexity), Nashua, NH, 2020

Abstract

Large-scale self-organizing systems often exhibit emergent behaviors which manifest as reduced-order dynamics on a low-dimensional manifold with a dimension much smaller than that of the original state-space. The ability to influence such self-organizing systems in a meaningful manner relies on observing the resulting emergent behaviors. While prior research has examined observability-related concepts from the viewpoint of network structure and connectivity in multiagent system, there exist limited insights into macroscopic-scale observability, i.e. the ability to observe reduced-order states of self-organizing systems. In this work, the ability to perceive or observe emergent behaviors in complex systems has been studied, and the relationship between an observability metric and model orders has been presented. Krylov subspace-based methods have been used to perform model order reduction for nonlinear systems such as coupled Rössler systems and interconnected electrical circuits that exhibit low-manifold emergent behaviors. The resulting numerical simulations indicate that reduced-order models, which are representative of emergent phenomena, usually possess higher observability metrics.

Near-optimal moving average estimation at characteristic timescales: An Allan variance approach

Hossein Haeri*, Craig Beal, and Kshitij Jerath
Journal PaperIEEE Control Systems Letters 5.5 (2020): 1531-1536

Abstract

A major challenge in moving average (MA) estimation is the selection of an appropriate averaging window length or timescale over which measurements remain relevant to the estimation task. Prior works typically perform timescale selection by examining multiple window lengths (or models) before selecting the `optimal' one using heuristics, domain knowledge expertise, goodness-of-fit, or information criterion (e.g., AIC, BIC etc.). In the presented work, we propose an alternative mechanism based on Allan Variance (AVAR) that obviates the need for assessing multiple models and systematically reduces reliance on heuristics or rules-of-thumb. The Allan Variance approach is used to identify the timescale that minimizes bias, thus determining the timescale over which past information remains most relevant. We also introduce an alternative method to obtain AVAR for unevenly spaced timeseries. The results from moving average estimation using an Allan Variance-determined window length are compared to the optimal moving average estimator that minimizes mean square error (MSE) for a variety of signals corrupted with Gaussian white noise. While the relevant timescales determined through AVAR tend to be longer than those associated with minimum MSE (i.e., AVAR-based MA estimation requires more measurements spread over a longer period of time), the AVAR-based moving average approach provides a valuable, systematic technique for near-optimal simple moving average estimation.

Thermodynamics-inspired Macroscopic States of Bounded Swarms

Hossein Haeri*, Jacob Leachman, and Kshitij Jerath
Journal Paper ASME Letters in Dynamic Systems and Control, vol. 1, pp 011-015, 2020

Abstract

The collective behavior of swarms is extremely difficult to estimate or predict, even when the local agent rules are known and simple. The presented work seeks to leverage the similarities between fluids and swarm systems to generate a thermodynamics-inspired characterization of the collective behavior of robotic swarms. While prior works have borrowed tools from fluid dynamics to design swarming behaviors, they have usually avoided the task of generating a fluids-inspired macroscopic state (or macrostate) description of the swarm. This work will bridge the gap by seeking to answer the following question: is it possible to generate a small set of thermodynamics-inspired macroscopic properties that may later be used to quantify all possible collective behaviors of swarm systems? In this paper, we present three macroscopic properties analogous to pressure, temperature, and density of a gas to describe the behavior of a swarm that is governed by only attractive and repulsive agent interactions. These properties are made to satisfy an equation similar to the ideal gas law and also generalized to satisfy the virial equation of state for real gases. Finally, we investigate how swarm specifications such as density and average agent velocity affect the system macrostate.

Observability Variation in Emergent Dynamics: A Study using Krylov Subspace-based Model Order Reduction

Zhaohui Yang* and Kshitij Jerath
Conference Paper 2020 American Control Conference (ACC), Denver, CO, USA, 2020, pp. 3461-3466.

Abstract

Large-scale self-organizing systems often exhibit emergent behaviors which manifest as reduced-order dynamics on a low-dimensional manifold with a dimension much smaller than that of the original state-space. The ability to influence such self-organizing systems in a meaningful manner relies on observing the resulting emergent behaviors. While prior research has examined observability-related concepts from the viewpoint of network structure and connectivity in multiagent system, there exist limited insights into macroscopic-scale observability, i.e. the ability to observe reduced-order states of self-organizing systems. In this work, the ability to perceive or observe emergent behaviors in complex systems has been studied, and the relationship between an observability metric and model orders has been presented. Krylov subspace-based methods have been used to perform model order reduction for nonlinear systems such as coupled Rössler systems and interconnected electrical circuits that exhibit low-manifold emergent behaviors. The resulting numerical simulations indicate that reduced-order models, which are representative of emergent phenomena, usually possess higher observability metrics.

Examining the Observability of Emergent Behavior as a Function of Reduced Model Order

Zhaohui Yang* and Kshitij Jerath
Conference Paper 2018 Annual American Control Conference (ACC), Milwaukee, WI, USA

Abstract

Current work on complex systems has been heavily focused on network interactions and network structure, without significant emphasis on the emergent phenomena that characterize such systems. In the included work, the ability to perceive or observe emergent behavior in complex systems has been studied. Specifically, by identifying emergent behavior as the dynamics of reduced-order models of the self-organizing systems, the observability of the emergent phenomena can be studied as a function of the model order. Included analytical results show the relationship between observability metrics of the full- and reduced-order models. Singular perturbation techniques have been used to perform model order reduction for nonlinear systems such as the Hyper Rössler and coupled Hindmarsh-Rose neuron models. Results indicate that accuracy increases and observability decreases with increasing model order. The trade-off between accuracy and observability metrics has been used to propose a predictive metric that identifies the desirable order at which to model emergent behavior.

Information-based Multi-robot Navigation, Exploration and Coverage using Adaptive Occupancy Grids

Mitchell Scott*
Thesis M.S., Mechanical Engineering, Washington State University, May 2018

Abstract

This thesis examines the control of an information-based multi-robot system (MRS). We demonstrate a simulated MRS exploration-coverage mission as the system converges to regions of high entropy (uncertainty), as it maximizes information content. Furthermore, the MRS mission utilizes an information-based adaptive occupancy grid which adaptively discretizes the grid so that map memory size is substantially reduced without sacrificing substantial map information content. We present a theoretical bounds on the information loss in the adaptive grid, while also demonstrating its ability to adaptively map an environment so that only high-entropy regions are finely discretized. Results indicate that the system can effectively explore an environment, converge to regions of high entropy, and further probe these regions to reduce the associated entropy (i.e. obtain more information), while utilizing a map that uses a fraction of the memory requirements but containing a comparable level of information. The novel contributions of this work include the development of an information-based adaptive grid and the leveraging of MRS control laws with an entropy-based world to control an MRS towards regions of interest.

Multi-robot Exploration and Coverage: Entropy-based Adaptive Maps with Adjacency Control Laws

Mitchell Scott* and Kshitij Jerath
Conference Paper 2018 Annual American Control Conference (ACC), Milwaukee, WI, USA

Abstract

Prior works have sought to improve exploration-coverage algorithms by developing multi-agent control techniques that either account for the states of neighboring agents, or utilize information-theoretic concepts to explore regions of high information content. In addition, variably-sized grid maps have been investigated to reduce computational expense. However, to the authors' knowledge, these three aspects have not been effectively harnessed together. In this paper, we present a multi-agent exploration and coverage mission which uses an entropy grid to effectively guide agents to explore and cover regions of high uncertainty, while also adaptively re-sizing and discretizing the grid world representation in real time. The adaptive grid presented is an improvement over constant-sized grids due to lower memory and processing requirements while still containing comparable information content. The entropy grid, when combined with adjacency-based control, allows the multi-agent system to effectively explore and cover environments that have time-varying target densities. The effectiveness of our approach is demonstrated through a real-time simulation with 75 agents.

Mission Performance Evaluation of Low-speed Small Unmanned Aerial Systems using Virtual Range and Stereo Camera Sensors

Mitchell Scott* and Kshitij Jerath
Conference Paper 2018 AIAA Information Systems-AIAA Infotech @ Aerospace, Kissimmee, FL, USA

Abstract

Current trends in miniaturization and the requirements of commercialization are resulting in an increased focus on small Unmanned Aerial Systems (sUAS). However, conceptual design and mission performance of these sUAS remains a costly endeavor requiring multiple iterations. In this paper, a virtual reality framework is used to assess the mission performance of a representative quadrotor. The mission profile considers waypoint navigation through a cluttered forest environment simulated using the MORSE-Blender platform. The performance assessment is carried out for three different range and three different stereo camera sensors with varying specifications. The flight time, power usage, and rate of virtual collisions are recorded during a Monte Carlo simulation to evaluate the mission performance. Hypotheses pertaining to the trade-offs between sensor specifications, mission flight time, distance traveled, collision likelihood, and mission completion are also examined. Results indicate that range sensors are more capable of avoiding obstacles and that two of the three range sensor configurations in the simulations traveled significantly farther than the remaining sensors.

Bridging the gap between sensor noise modeling and sensor characterization

Kshitij Jerath, Sean Brennan, and Constantino Lagoa
Journal PaperMeasurement, vol. 116, pp. 350-366 (2018)

Abstract

In prior works, the tasks of noise modeling and sensor characterization have typically been studied independently of each other. In spite of extensive research on noise modeling and sensor characterization, there still exists a need to bridge the gap between parameters used in sensor noise models, and quantities used to characterize commercially available sensors. The included work addresses this need by presenting tutorial-style exemplary analyses that relate noise model parameters to sensor characteristics for some common noise types found in sensors. Specifically, this work seeks to demonstrate that sensor noise characterization techniques can be applied to simulated sensor noise data to recover the original noise model parameters. The presented relationships between noise models and sensor characterization tools can help engineers and scientists numerically verify the expected performance bounds of their systems using simulated signals from commercially available sensors. Moreover, these numerical tools can also guide design engineers towards developing or selecting sensors with specifications especially suited to the design constraints of a given application.

Cooperative Adaptive Cruise Control: Impact on Self-organized Traffic Jams

Taehooie Kim*
Thesis M.S., Mechanical Engineering, Washington State University, May 2017

Abstract

This thesis addresses the impact of Cooperative Adaptive Cruise Control (CACC) on the formation of self-organized traffic jams. Self-organized traffic jams are also known as phantom jams and appear on a road when the vehicular density increases and exceeds a certain threshold (i.e., critical density). The master equation is used to model a self-organized traffic jam at mesoscopic scale. Furthermore, to examine the vehicular behavior under the traffic condition, the master equation utilizes the General Motor(GM)-fourth car-following model to obtain transition probability rates. Since CACC-enabled vehicle can obtain additional traffic states (including congestion information) via Dedicated Short Range Communication (DSRC), the existing car-following model is to be modified for deriving the behavior of CACC-enabled vehicles dependent on both traffic environment and vehicle states. By using the new CACC algorithm, this paper examines the role of CACC-enabled vehicles in mixed environment on traffic flow. As the proportion of CACC-enabled vehicles increases, it is shown that these changes lead to an increase in traffic mobility while maintaining low susceptibility and mitigating the rate of forming a cluster or jam.

Origins of Collective Intelligence: Is there a Homology between Social Networks and Knowledge Structures?

Charles Pezeshki and Kshitij Jerath
Abstract Collective Intelligence Conference, Jun 2017, New York, USA

Mitigation of self-organized traffic jams using cooperative adaptive cruise control

Taehooie Kim* and Kshitij Jerath
Conference Paper International Conference on Connected Vehicles and Expo (ICCVE), Sep 2016, Seattle, WA, USA

Abstract

Prior work has shown that while vehicles equipped with Adaptive Cruise Control (ACC) algorithms have the ability to improve traffic flows by increasing the critical density at which traffic jams begin to occur, they remain highly susceptible to the presence of even a few human-driven vehicles. This necessitates a trade-off assessment between improving traffic flow and reducing its susceptibility to presence of human-driven vehicles. In this paper, we address this issue via Cooperative Adaptive Cruise Control (CACC) algorithms that use information communicated via other vehicles or infrastructure. Specifically, the CACC algorithms use a modified form of the General Motors' car-following model, where the driver sensitivity is a function of the size of the existing traffic jam. Analysis using the master equation approach shows that, by altering its response based on the existing traffic state, the following connected vehicle is able to mitigate jam formation for a wide range of vehicular densities. Moreover, since the following connected vehicle can leverage information across larger distances, the `effective' critical density is higher than previous results developed with ACC-enabled vehicles only. The included results indicate that traffic state-dependent CACC algorithms can improve traffic flow, and also hold significant potential to reduce susceptibility to the presence of human-driven vehicles.

Simulation framework for incorporating sensor systems in UAS conceptual design

Kshitij Jerath and Jack W. Langelaan
Conference Paper AIAA Science and Technology Forum and Exposition (SciTech 2016), San Diego, CA, USA

Abstract

Sensor systems typically have not played a significant role in the aircraft conceptual design process. However, with the proliferation of unmanned aerial systems (UAS), the role of sensor systems on vehicle conceptual design can no longer be ignored. This issue is accentuated at smaller scales, where the sensing equipment may place significant constraints on the size, weight and power requirements of the aircraft. In this paper, we present a novel method that incorporates sensor systems into vehicle conceptual design and mission capability analysis. The presented method relies on a simulation framework in which the Blender Game Engine is used to generate complex cluttered flight environments. The MORSE simulator is used to fly vehicles with varying sensor configurations in these cluttered environments in order to assess mission capabilities. Results show the applicability of the presented approach for assessing waypoint navigation mission capabilities as a function of sensor configurations and specifications.

Dynamic Prediction of Vehicle Cluster Distribution in Mixed Traffic: A Statistical Mechanics-Inspired Method

Kshitij Jerath, Asok Ray, Sean N. Brennan, and Vikash V. Gayah
Journal Paper IEEE Trans. on Intelligent Transportation Systems (in press), 2015

Abstract

The advent of intelligent vehicle technologies holds significant potential to alter the dynamics of traffic flow. Prior work on the effects of such technologies on the formation of self-organized traffic jams has led to analytical solutions and numerical simulations at the mesoscopic scale, which may not yield significant information about the distribution of vehicle cluster size. Since the absence of large clusters could be offset by the presence of several smaller clusters, the distribution of cluster sizes can be as important as the presence or absence of clusters. To obtain a prediction of vehicle cluster distribution, the included work presents a statistical mechanics-inspired method of simulating traffic flow at a microscopic scale via the generalized Ising model. The results of the microscopic simulations indicate that traffic systems dominated by adaptive cruise control (acc)-enabled vehicles exhibit a higher probability of formation of moderately sized clusters, as compared with the traffic systems dominated by human-driven vehicles; however, the trend is reversed for the formation of large-sized clusters. These qualitative results hold significance for algorithm design and traffic control because it is easier to predict and take countermeasures for fewer large localized clusters as opposed to several smaller clusters spread across different locations on a highway.

Identification of locally influential agents in self-organizing multi-agent systems

Kshitij Jerath and Sean N. Brennan
Conference Paper Proceedings of the American Control Conference, July 2015, Chicago, IL, Pages 335-340

Abstract

Current research methods directed towards measuring the influence of specific agents on the dynamics of a large-scale multi-agent system (MAS) rely largely on the notion of controllability of the full-order system, or on the comparison of agent dynamics with a user-defined macroscopic system property. However, it is known that several large-scale multiagent systems tend to self-organize, and their dynamics often reside on a low-dimensional manifold. The proposed framework uses this fact to measure an agent’s influence on the macroscopic dynamics. First, the minimum embedding dimension that can encapsulate the low-dimensional manifold associated with the self-organized dynamics is identified using a modification of the method of false neighbors. Second, the minimum embedding dimension is used to guide the Krylov subspace-based model order reduction of the system dynamics. Finally, an existing controllability-based metric is applied to the local reducedorder representation to measure an agent’s influence on the self-organized dynamics. With this technique, one can identify regions of the state space where an agent has significant local influence on the dynamics of the self-organizing MAS. The proposed technique is demonstrated by applying it to the problem of vehicle cluster formation in traffic, a prototypical self-organizing system. As a result, it is now possible to identify regions of the roadway where an individual driver has the ability to influence the dynamics of a self-organized traffic jam.

Influential subspaces in self-organizing multi-agent systems

Kshitij Jerath
Dissertation Ph.D. Dissertation, Mechanical Engineering, The Pennsylvania State University, December 2014

Abstract

This dissertation addresses the issue of influence in self-organizing multi-agent systems by using traffic jams as a prototypical example of self-organized behavior. Specifically, the problem of ascertaining the influence of a set of agents on the ensemble dynamics is addressed through two complementary approaches. In the first approach, discussed in Part I of the dissertation, the ability to influence ensemble dynamics is studied as a function of changing agent population demographics. Statistical mechanics-inspired methodologies, such as the master equation and the generalized Ising model, are used to study the effect of introduction of vehicles equipped with adaptive cruise control (ACC) algorithms on the self-organized dynamics of traffic jams. Results indicate mixed positive and negative effects of introduction of acc-equipped vehicles at various traffic densities. While this approach can help guide long-term intelligent vehicle deployment strategies on the time scale of years or decades, population demographic control is not a feasible solution for influencing large-scale multi-agent systems on the time scale of minutes or hours. Thus, the second approach, discussed in Part II of this dissertation, addresses the problem by identifying appropriate regions of the state space within which the control efforts exerted by a small set of agents can influence the self-organized dynamics of the ensemble. The methodologies adopted in this approach make use of the kinematic wave theory of traffic flow and the notion of controllability to present the novel concept of influential subspaces. Results indicate that there exists a strong spatial dependence that governs an agent’s ability to influence the self-organized dynamics of large-scale multi-agent systems.

Next-generation Vehicle Positioning Techniques for GPS-Degraded Environments to Support Vehicle Safety and Automation Systems

Auburn University, Stanford Research Institute, The Pennsylvania State University, and Kapsch TrafficCom Inc.
Techincal Report Federal Highway Administration (FHWA)

Abstract

The next generation of safety and vehicle automation will rely on precise positioning, yet GPS-based positioning is hampered by blockages of the global positioning system (GPS) signal—a broader approach is needed that does not rely exclusively on GPS. However, there is no one "silver bullet"; therefore, this proposed work seeks to achieve a major improvement in vehicle positioning performance by developing a multifaceted approach to achieving precise positioning, even in situations in which the GPS signal is inadequate due to a dense tree canopy, building shadowing, or other factors. The work builds on previous work done by the partners in other domains to apply these techniques to future safety and automation applications. Three key technology areas hold promise individually, and the research intends to show that a combined, integrated system is even more powerful by exploiting the strengths of each technique. First, terrain-based localization (based on precise measurements of vehicle pitch and roll, combined with wheel odometry) can be readily used to find the vehicle's absolute longitudinal position within a premapped highway segment, compensating for drift, which occurs in dead-reckoning systems in long longitudinal stretches of road. Second, visual odometry keys upon visual landmarks at a detailed level to correlate position to a (visually) premapped road segment to find vehicle position along the roadway. Both of these preceding techniques rely on foreknowledge of road features; in essence, a feature-enhanced version of a digital map. This becomes feasible in the "connected vehicle" fixture, in which tomorrow's vehicles have access to quantities of data orders of magnitude greater than today's cars, as well as the ability to share data at high data rates. The third technology approach relies on radio frequency ranging based on dedicated short range xommunications (DSRC) radio technology. In addition to pure radio frequency ranging with no GPS signals, information from radio frequency ranging can be combined with GPS-range measurements (which may be inadequate on their own) to generate a useful position. (2014)

Influential subspaces of connected vehicles in highway traffic

Kshitij Jerath, Vikash V. Gayah, and Sean N. Brennan
Conference Paper Symposium Celebrating 50 Years of Traffic Flow Theory, August 2014, Portland, OR

Abstract

This work introduces the novel concept of an influential subspace, with focus on its application to highway traffic containing connected vehicles. In this context, an influential subspace of a connected vehicle is defined as the region of a highway where the connected vehicle has the ability to positively influence the macrostate, i.e. the traffic jam, so as to dissipate it within a specified time interval. Analytical expressions for the influential subspace are derived using the LighthillWhitham-Richards theory of traffic flow. Included results describe the span of the influential subspace for specific traffic flow conditions and pre-specified driving algorithms of the connected vehicles.

Statistical mechanics-inspired framework for studying the effects of mixed traffic flows on highway congestion

Kshitij Jerath, Asok Ray, Sean N. Brennan, and Vikash V. Gayah
Conference Paper Proceedings of the American Control Conference, June 2014, Portland, OR, Pages 5402-5407

Abstract

Intelligent vehicles equipped with adaptive cruise control (ACC) technology have the potential to significantly impact the traffic flow dynamics on highways. Prior work in this area has sought to understand the impact of intelligent vehicle technologies on traffic flow by making use of mesoscopic modeling that yields closed-form solutions. However, this approach does not take into account the self-organization of vehicles into clusters of different sizes. Consequently, the predicted absence of a large traffic jam might be inadvertently offset by the presence of many smaller clusters of jammed vehicles. This study - inspired by research in the domain of statistical mechanics - uses a modification of the Potts model to study cluster formation in mixed traffic flows that include both human-driven and ACC-enabled vehicles. Specifically, the evolution of self-organized traffic jams is modeled as a non-equilibrium process in the presence of an external field and with repulsive interactions between vehicles. Monte Carlo simulations of this model at high vehicle densities suggest that traffic streams with low ACC penetration rates are likely to result in larger clusters. Vehicles spend significantly more time inside each cluster for low ACC penetration rates, as compared to streams with high ACC penetration rates.

Human detection to increase safety in complex construction environments

Sean Brennan, Pramod Vemulapalli, Kshitij Jerath, Michael Robinson, and M. Guo
Technical Report Volvo Construction Equipment (2012)

Abstract

This report documents the need and motivation of the project, provides a brief overview of the literature on LIDAR-based object detection and tracking algorithms, and lists the algorithms investigated and developed by Penn State. Additionally, the report also documents results and conclusions derived from data collected during the site visit to Volvo Construction Equipment, Shippensburg, PA. These results indicate that the LIDAR equipment can function safely on the compactor even when the drum vibrations are turned on. The current build of the rule-based algorithm developed at Penn State is able to correctly identify foreground objects from the background approximately 90% of the time. The report also includes a prototype of an advanced object detection and tracking algorithm that could potentially overcome the limitations of rule-based algorithms.

Analytical prediction of self-organized traffic jams as a function of increasing ACC penetration

Kshitij Jerath and Sean N. Brennan
Journal Paper IEEE Trans. on Intelligent Transportation Systems, Volume 13, Issue 4, 2012, Pages 1782-1791

Abstract

Self-organizing traffic jams are known to occur in medium-to-high density traffic flows, and it is suspected that adaptive cruise control (ACC) may affect their onset in mixed human-ACC traffic. Unfortunately, closed-form solutions that predict the occurrence of these jams in mixed human-ACC traffic do not exist. In this paper, both human and ACC driving behaviors are modeled using the General Motors fourth car-following model and are distinguished by using different model parameter values. A closed-form solution that explains the impact of ACC on congestion due to the formation of self-organized traffic jams (or “phantom” jams) is presented. The solution approach utilizes the master equation for modeling the self-organizing behavior of traffic flow at a mesoscopic scale and the General Motors fourth car-following model for describing the driver behavior at the microscopic scale. It is found that, although the introduction of ACC-enabled vehicles into the traffic stream may produce higher traffic flows, it also results in disproportionately higher susceptibility of the traffic flow to congestion.

GPS-free terrain-based vehicle tracking on road networks

Kshitij Jerath, and Sean N. Brennan
Conference Paper Proceedings of the American Control Conference, June 2012, Montreal, QC

Abstract

Prior experiments have confirmed that specific GPS-free terrain-based localization algorithms can perform vehicle tracking in real-time on a single road segment at a time. However, the ability of these algorithms to perform vehicle tracking on large road networks, i.e. across intersections and multiple road segments, has not been verified. In this study, it is shown that it is possible to build upon the existing terrain-based localization algorithms to maintain vehicle tracking in large road networks. A set of estimators based on the Unscented Kalman Filter framework is used to track the vehicle in a section of a road network, i.e. across a few road segments and an intersection. A multiple model estimation scheme, based on comparing incoming attitude measurements with a terrain map, is used to identify the road segment that the vehicle is currently traveling over. Experiments indicate that it is possible to maintain vehicle tracking as a vehicle travels across an intersection in a road network.

GPS-free terrain-based vehicle tracking performance as a function of inertial sensor characteristics

Kshitij Jerath, and Sean N. Brennan
Conference Paper Proceedings of the Dynamics Systems and Control Conference, Oct 2011, Arlington, VA

Abstract

Prior experiments have confirmed that specific terrain-based localization algorithms, designed to work in GPS-free or degraded-GPS environments, achieve vehicle tracking with tactical-grade inertial sensors. However, the vehicle tracking performance of these algorithms using low-cost inertial sensors with inferior specifications has not been verified. The included work identifies, through simulations, the effect of inertial sensor characteristics on vehicle tracking accuracy when using a specific terrain-based tracking algorithm based on Unscented Kalman Filters. Results indicate that vehicle tracking is achievable even when low-cost inertial sensors with inferior specifications are used. However, the precision of vehicle tracking decreases approximately linearly as bias instability and angle random walk coefficients increase. The results also indicate that as sensor cost increases, the variance in vehicle tracking error asymptotically tends to zero. Put simply, as desired precision increases, increasingly larger and quantifiable investment is required to attain an improvement in vehicle tracking precision.

A failure rate analysis of complex vehicles

Joseph Yutko, Kshitij Jerath, and Sean N. Brennan
Journal Paper International Journal of Heavy Vehicle Systems, Volume 17, Issue 1, 2010, Pages 76-98

Abstract

When engineered items fail, there are often indicators of decay long before the system collapses. This research explores this concept applied to complex vehicles operated in public transportation, and the results can be extrapolated to any vehicle system. Transit bus reliability data gathered from eight transit agencies distributed across the United States are analyzed at a vehicle and subsystem level to identify system failures. An analysis of vehicle subsystem component failures is conducted, where the theory of reliability of repairable systems is applied to the in-transit data to determine if major component failures can be detected by increases in cumulative and subsystem failure rates. Thus, the impact of the research illustrates that major repairs might be detected far in advance of when they are needed.

Adaptive Cruise Control: Towards higher traffic flows at the cost of increased susceptibility to congestion

Kshitij Jerath, and Sean N. Brennan
Conference Paper 10th International Symposium on Advanced Vehicle Control, August 2010, Loughborough, UK

Abstract

Self-organizing traffic jams are known to occur in medium-to-high density traffic flows and it is suspected that ACC algorithms may affect their onset in mixed human-ACC traffic flows. Unfortunately, closed-form solutions that predict the statistical occurrence of these jams in mixed traffic do not exist. In this paper, a closed form solution that explains the impact of adaptive cruise control (ACC) on congestion due to the formation of self-organizing traffic jams (or “phantom” jams) is obtained. The master equation approach is selected for developing a model that describes the self-organizing behavior of traffic flow at a mesoscopic scale. The master equation approach is further developed to incorporate driver (or agent) behavior using ACC or car-following algorithms. The behavior for both human-driven and ACC vehicles is modeled using the General Motors‟ fourth model. It is found that while introduction of ACC vehicles into traffic may enable higher traffic flows, it also results in disproportionately higher susceptibility of the traffic flow to congestion.

Impact of Adaptive Cruise Control on the formation of self-organized traffic jams on highways

Kshitij Jerath
Thesis M.S., Mechanical Engineering, The Pennsylvania State University, May 2010

Abstract

This thesis describes the analysis of a potential control mechanism for self-organizing systems by studying the specific problem of self-organizing traffic jams. Self-organizing traffic jams are known to occur in medium-to-high density traffic flows. Various techniques for modeling traffic flow are discussed and their advantages and limitations are considered. The master equation approach is selected for developing a model that describes the self-organizing behavior of traffic flow at a mesoscopic scale. The master equation approach is further developed to incorporate driver (or agent) behavior. Control of the self-organizing system is presented via introduction of similar agents with slightly varying interaction properties. The introduction of such agents into a self-organizing system is considered to be analogous to the introduction of vehicles with adaptive cruise control (ACC) into traffic flow. The behavior for both human-driven and ACC vehicles is modeled using the same driver model but with slightly different model parameters. It is found that introduction of a small percentage of agents with slightly different interaction behavior has the potential to affect the dynamics of the self-organizing system. Specifically, it is found that while introduction of ACC vehicles into traffic may enable higher traffic flows, it also results in disproportionately higher susceptibility of the traffic flow to congestion.

Study of the relationship between results of the Bus Testing Program and in-service performance of buses

Sean Brennan, Kshitij Jerath, David Klinikowski, Saravanan Muthiah, and Joseph Yutko
Technical Report Federal Transit Administration (FTA) (2008)

Abstract

When engineered items fail, there are often indicators of decay long before the system collapses. This research explores this concept applied to complex vehicles operated in public transportation, and the results can be extrapolated to any vehicle system. Transit bus reliability data gathered from eight transit agencies distributed across the United States are analyzed at a vehicle and subsystem level to identify system failures. An analysis of vehicle subsystem component failures is conducted, where the theory of reliability of repairable systems is applied to the in-transit data to determine if major component failures can be detected by increases in cumulative and subsystem failure rates. Thus, the impact of the research illustrates that major repairs might be detected far in advance of when they are needed.

Intellectual Property Monetization by R&D Organizations in India and China

R. Deshpande, D. Johar, A. Kasyap, C. Feng, K. Jerath, Z. Li
Conference Paper Fifth International Symposium on Management of Technology, Hangzhoue, China, pp. 681-683 (2007)