Aron Laszka |

SAGE: Similarity-Based Cleaning of Poisoned Training Data from Verified Examples

Sun, 01 Nov 2026 00:00:00 +0000

Control Barrier Function Based Speed Control for Fixed-Line Transit Headway Management

Tue, 01 Sep 2026 00:00:00 +0000

Column Generation for the Micro-Transit Zoning Problem

Sat, 01 Aug 2026 00:00:00 +0000

Computing Headway Bounds Under Worst-Case Bunching in Fixed-Line Transit Systems

Sat, 01 Aug 2026 00:00:00 +0000

NSF CAREER: Scaling Deep Reinforcement Learning for Societal-Scale Cyber-Physical Systems

Wed, 01 Jul 2026 00:00:00 +0000

This project will develop artificial intelligence techniques to improve the operation of large-scale infrastructure systems, such as smart transportation networks and electric power grids, which are essential to modern life. These complex systems, known as societal-scale cyber-physical systems, integrate physical infrastructure with thousands of sensors, computing devices, and actuators. Due to their scale and distributed nature, managing these systems in real time poses a significant challenge. To address this challenge, the project will leverage deep reinforcement learning, a form of artificial intelligence that learns optimal decision-making strategies directly from data. By improving the efficiency and reliability of critical infrastructure systems, the research will further the national interest through reducing traffic congestion, improving emergency response times, and increasing the stability of the power grid. This project will also develop a highly skilled science, technology, engineering, and mathematics (STEM) talent pipeline. These development activities will include creating interactive, game-based learning modules for K-12 students, developing an interdisciplinary graduate course on AI and cyber-physical systems, and providing training materials to help community partners effectively adopt artificial intelligence technologies.

This project will advance the foundations of sequential decision-making for societal-scale cyber-physical systems (CPS) by addressing the fundamental scalability and communication challenges faced by existing deep reinforcement learning (DRL) approaches. The scope of the research will encompass theoretical and algorithmic innovations to develop a hierarchical and communication-aware DRL framework, which will be tailored to the vast state-action spaces and distributed nature of societal-scale CPS. First, the project will introduce a cyber-physical feudal reinforcement learning framework that will provide automated, data-driven system decomposition. This approach will partition a complex cyber-physical system into tractable subsystems based on their underlying dynamic cyber-physical couplings. The framework will then train a hierarchy of DRL policies to control the system: a high-level coordination policy will generate abstract but physically achievable goals for ensembles of decentralized, low-level subsystem policies. Second, to address the communication constraints of distributed CPS, the project will extend the feudal DRL framework to be communication-aware. To minimize bandwidth overhead, the framework will integrate physics-informed causal abstraction to compress state representations and trigger communication only when necessary. To handle communication disruptions and missing data, the framework will employ adversarial curriculum learning to train robust policies that will rely on belief states and learned intrinsic motivation to ensure graceful degradation. The project will evaluate the proposed innovations using high-fidelity simulations, real-world datasets from transportation and power systems, and community partnerships, rigorously demonstrating the intellectual contributions required to deploy scalable, data-driven control policies across critical distributed infrastructure.

Adversarial Reinforcement Learning for Detecting False Data Injection Attacks in Vehicular Routing

Fri, 01 May 2026 00:00:00 +0000

Dynamic Vehicle Routing Problem with Prompt Confirmation of Advance Requests

Fri, 01 May 2026 00:00:00 +0000

MoveOD: Synthesizing Origin-Destination Commute Distribution from U.S. Census Data

Fri, 01 May 2026 00:00:00 +0000

Ensemble Prediction of Task Affinity for Efficient Multi-Task Learning

Wed, 01 Apr 2026 00:00:00 +0000

AI for Smart Transportation

Thu, 01 Jan 2026 00:00:00 +0000

An End-to-End Solution for Public Transit Stationing and Dispatch Problem

Sat, 01 Nov 2025 00:00:00 +0000

Fixed-Line Transit 2.0: Real-Time Optimization of High-Frequency Transit Service

Tue, 01 Jul 2025 00:00:00 +0000

This project will develop and deploy a real-time, AI-driven decision support system that makes high-frequency bus service more reliable by anticipating and proactively mitigating delays, overcrowding, and service disruptions. Frequent, high-ridership bus routes are inherently unstable: when a bus falls behind, it encounters more waiting passengers and falls even further behind, until buses bunch together and leave long gaps in service, causing longer and less predictable waits, overcrowding, and missed transfers. The project focuses on WeGo Public Transit’s Route 55 Murfreesboro in Nashville, Tennessee, WeGo’s busiest route with nearly 5,000 weekday boardings, but also its least reliable one: more than 1 in 4 buses are either bunched or gapped, adding up to over 105 hours of excess waiting time for passengers every weekday. The project is led by WeGo in partnership with the Nashville Department of Transportation and Multimodal Infrastructure, the Pennsylvania State University, Vanderbilt University, and technology companies.

At the heart of the project is an Integrated Decision Support System that fuses heterogeneous real-time data, such as bus locations updated every two seconds, passenger counts, traffic, and weather. Computer vision will count passengers at bus stops and on buses, and machine learning models will forecast bus arrival times, headways, overcrowding, and service disruptions. Building on these forecasts, the system will warn dispatchers of emerging problems and use AI-driven optimization to recommend strategies that prevent them, such as holding or short-turning buses, skipping stops, or inserting standby buses. Adaptive traffic signal control with headway-conditional transit signal priority, continuously optimized using machine learning, will further reduce travel times and their variability, and riders will receive more accurate arrival predictions in their smartphone apps and on digital displays at bus stops.

Overview of the Fixed-Line Transit 2.0 system

Together with Vanderbilt University, our team at Penn State is responsible for the AI at the core of this system: prediction algorithms for bus arrivals and headways, and a real-time decision support system for dispatchers and drivers. We will create a decision-focused learning framework, in which prediction models trained on spatiotemporal data, such as weather, traffic, and ridership, are incorporated into non-myopic online optimization algorithms for proactive headway management and vehicle stationing. These algorithms will use the learned models to simulate uncertain future trajectories of the transit system and reason through them to improve service reliability. Our approach will combine these strengths of AI with the strengths of human supervisors, whose tacit knowledge and expertise are invaluable for understanding the dynamic challenges of fixed-line transit, while giving supervisors, drivers, and passengers real-time situational awareness of the whole system.

By making frequent bus service faster and more predictable, the project will improve access to jobs, education, and essential services: 65% of Route 55 riders do not have access to a vehicle, and since the route has more transfers to and from other routes than any other WeGo route, the benefits will extend to residents in many other parts of Nashville. More reliable transit may also draw people away from driving, and adaptive signals with transit signal priority will reduce delays at intersections, together reducing congestion and improving energy efficiency. The decision support system will also improve the productivity and job quality of dispatchers and drivers. Unlike dedicated transit lanes, whose construction is a multi-year, resource-intensive undertaking, the system builds on existing infrastructure, such as municipal fiber, onboard sensors, and real-time transit data feeds, greatly reducing the cost, duration, and risk of deployment. Extending it to six additional high-frequency WeGo routes would bring more than 2 in 3 WeGo bus boardings, nearly 20,000 rides per weekday, onto “Transit 2.0” routes, and the approach is relevant to other transit agencies facing similar challenges.

Scalable Decision-Making in Stochastic Environments through Learned Temporal Abstraction

Tue, 01 Apr 2025 00:00:00 +0000

AI-Powered Autonomy-Aware Neighborhood Mobility Zones

Tue, 01 Oct 2024 00:00:00 +0000

This project reimagines public transportation for mid-sized U.S. cities through an AI-driven, multi-modal mobility framework that integrates fixed-line buses and on-demand microtransit into a single, responsive system. Led by the Chattanooga Area Regional Transportation Authority (CARTA) with technical contribution from the Pennsylvania State University, Vanderbilt University, and Cornell University, this $7 million initiative is designed to enhance infrastructure utilization, mobility access, and financial viability of public transit in Chattanooga while providing a national model for AI-enabled mobility resilience.

At the project’s core is the concept of neighborhood mobility zones—virtual, data-driven operational areas that dynamically organize and optimize transportation resources without requiring costly new infrastructure. These zones centralize around Chattanooga’s downtown core, connecting neighborhoods, universities, hospitals, and economic centers. The system uses AI algorithms for real-time multi-modal routing, predictive demand modeling, and autonomy-aware fleet planning, ensuring that resources, human-driven and autonomous, are deployed where they are most effective.

Beyond technical innovation, the project establishes sustainable business models, such as subscription-based mobility packages and monetization of CARTA’s 4,200 parking spaces and other infrastructures, to ensure long-term economic viability.

By integrating multimodal mobility options, along with personal cars, the goal is to increase the overall trips served through the system by 5x. The project advances DOE’s Energy Efficient Mobility Systems (EEMS) objectives while aligning with the March 2025 Executive Order on “Achieving Efficiency Through State and Local Preparedness.” Overall it demonstrates how neurosymbolic AI and integrating autonomous technology with multimodal options can improve infrastructure reliability and effectiveness.

Rural Transit Solutions for Enhancing Access & Connectivity

Tue, 01 Oct 2024 00:00:00 +0000

This Civic Innovation Challenge (CIVIC) Stage 1 project will perform research to address transportation challenges faced by residents and workers in rural areas around industrial hubs. Rapid economic growth in these areas demands efficient public transit systems that can serve a geographically dispersed workforce with strict arrival times. However, existing solutions often struggle with these unique requirements, leading to traffic congestion, pollution, and limited access to essential services for residents. This research project will develop and deploy a novel multi-modal transit system for Blue Oval City (BOC), a new rural industrial hub in Stanton, Tennessee. The system will combine fixed-line buses with on-demand micro-transit services, addressing the challenges posed by the geographically dispersed workforce and strict arrival times. Further, the project will design equitable fare structures to ensure that the system is sustainable for all stakeholders. If successful, the research will benefit communities around BOC by enhancing their access to employment opportunities, essential resources, and services, ultimately improving quality of life and community resilience. Furthermore, developing efficient and reliable transit can promote the growth of businesses in the surrounding areas and offer more affordable living options. The research findings and deliverables will be transferable to other rural industrial hubs across the nation, fostering economic development and improving public transportation infrastructure.

This project will be led by a multi-disciplinary, multi-stakeholder team of researchers and civic partners to co-create a scalable pilot project for BOC. The research project has two main goals. (i) Develop a system to optimize multi-modal transit networks, including micro-transit services that act as feeders for large, fixed-line buses. Models based on artificial intelligence will be designed to handle high volumes of employee trips and create service zones with optimal pick-up and drop-off locations, guaranteeing on-time arrivals for workers regardless of their shift schedules. (ii) Explore sustainable funding plans through stakeholder engagement, designing a fare structure that incentivizes ridership and ensures social equity. This includes examining how the system can improve access to healthcare, education, and other essential services.

Grid-Aware Charging and Operational Optimization for Mixed-Fleet Public Transit

Sun, 01 Sep 2024 00:00:00 +0000

Deploying Mobility-On-Demand for All by Optimizing Paratransit Services

Thu, 01 Aug 2024 00:00:00 +0000

Acceptance Rate: 20%
Note: Multi-Year Track on AI and Social Good

SmartTransit.AI: A Dynamic Paratransit and Microtransit Application

Thu, 01 Aug 2024 00:00:00 +0000

Multi-Agent Reinforcement Learning with Hierarchical Coordination for Emergency Responder Stationing

Mon, 01 Jul 2024 00:00:00 +0000

An Online Approach to Solving Public Transit Stationing and Dispatch Problem

Wed, 01 May 2024 00:00:00 +0000

Acceptance Rate: 23%

Forecasting and Mitigating Disruptions in Public Bus Transit Services

Wed, 01 May 2024 00:00:00 +0000

Acceptance Rate: 25%

Multi-Agent Reinforcement Learning for Assessing False-Data Injection Attacks on Transportation Networks

Wed, 01 May 2024 00:00:00 +0000

Acceptance Rate: 25%

Grid Resilience Against Wildfire with Machine Learning: Machine Learning based Detection, Localization and Mitigation of the Impact of Forest Fires on Power Grids

Mon, 01 Jan 2024 00:00:00 +0000

Reinforcement Learning based Proactive Control for Enabling Power Grid Resilience to Wildfire

Mon, 01 Jan 2024 00:00:00 +0000

Impact Factor: 11.7

PhobosBC: A Blockchain-based Crowdsourced Post-disaster Mapping System and its Agent-based Simulation

Wed, 01 Nov 2023 00:00:00 +0000

Bug Hunters’ Perspectives on the Challenges and Benefits of the Bug Bounty Ecosystem

Tue, 01 Aug 2023 00:00:00 +0000

Correct-by-Design Interacting Smart Contracts and a Systematic Approach for Verifying ERC20 and ERC721 Contracts with VeriSolid

Sat, 01 Jul 2023 00:00:00 +0000

Impact Factor: 7

Survey and Taxonomy of Adversarial Reconnaissance Techniques

Thu, 01 Jun 2023 00:00:00 +0000

Impact Factor: 23.8

The Benefits of Vulnerability Discovery and Bug Bounty Programs: Case Studies of Chromium and Firefox

Mon, 01 May 2023 00:00:00 +0000

Acceptance Rate: 19%

Bayesian Models for Node-Based Inference Techniques

Sat, 01 Apr 2023 00:00:00 +0000

Impact of COVID-19 on Public Transit Accessibility and Ridership

Sat, 01 Apr 2023 00:00:00 +0000

Principled Data-Driven Decision Support for Cyber-Forensic Investigations

Wed, 01 Feb 2023 00:00:00 +0000

Acceptance Rate: 19.6%

Rolling Horizon based Temporal Decomposition for the Offline Pickup and Delivery Problem with Time Windows

Wed, 01 Feb 2023 00:00:00 +0000

Acceptance Rate: 19.6%

Strategic Cyber Camouflage

Sun, 01 Jan 2023 00:00:00 +0000

Decentralized Computation Market for Stream Processing Applications

Thu, 01 Sep 2022 00:00:00 +0000

Offline Vehicle Routing Problem with Online Bookings: A Novel Problem Formulation with Applications to Paratransit

Fri, 01 Jul 2022 00:00:00 +0000

Acceptance Rate: 14.9%

Neural Architecture and Feature Search for Predicting the Ridership of Public Transportation Routes

Wed, 01 Jun 2022 00:00:00 +0000

A Taxonomy of Blockchain Oracles: The Truth Depends on the Question

Sun, 01 May 2022 00:00:00 +0000

Acceptance Rate: 18.6%

An Online Approach to Solve the Dynamic Vehicle Routing Problem with Stochastic Trip Requests for Paratransit Services

Sun, 01 May 2022 00:00:00 +0000

Data-Driven Prediction and Optimization of Energy Use for Transit Fleets of Electric and ICE Vehicles

Tue, 01 Feb 2022 00:00:00 +0000

Strategic Remote Attestation: Testbed for Internet-of-Things Devices and Stackelberg Security Game for Optimal Strategies

Fri, 01 Oct 2021 00:00:00 +0000

AI for Equitable and Resilient Food Distribution during Disasters

Wed, 01 Sep 2021 00:00:00 +0000

Food insecurity is the lack of consistent and reliable access to nutritious food. In the Houston (Texas) area, over 14% of Harris County households experienced food insecurity before the emergence of COVID-19. It is unclear how the nutritional needs of Houston’s vulnerable populations will be served when the next devastating weather event strikes the region, given that the city is already experiencing multiple disasters, including COVID-19, economic disruptions, and systemic food insecurity. The Houston Food Bank (HFB) collaborates with over 1,500 partners to address the needs of families experiencing food insecurity. The COVID-19 pandemic has stretched funding and personnel significantly while driving substantial increases in demand. HFB has adapted to past natural disasters, such as hurricanes and floods, in an ad hoc manner and seeks to build disaster resilience by incorporating strategies beyond current practices. The urgent needs are to identify food-insecure communities, understand their nutritional needs during disasters, and provide nutritious and culturally appropriate food in a manner that preserves privacy and dignity. Left unaddressed, these issues may create failures in HFB’s ability to distribute food effectively, equitably, and efficiently.

Through a community-driven approach, this project brings together civic collaborators with university researchers to align HFB’s food distribution strategy to match food insecurity during multiple disaster profiles. Collaborators that have driven this approach include food pantries associated with HFB, local government, state government, other food banks, Foodbank Associations, workforce development and education, housing, health organizations, United Way, and YMCA. Leveraging lessons learned from six workshops conducted with HFB and its partners during our Stage 1 planning grant, we will jointly undertake a research-centered, socio-technical approach to develop and evaluate decision-making tools that incorporate equity indices for selecting food distribution hubs to improve the resilience of the food distribution network. Our project has the following objectives:

develop indicators of individual and community equity for food distribution during pandemics and extreme weather events;
design a network organizational resilience index for food bank networks and interventions to improve network resilience;
develop and validate a predictive tool for infrastructure vulnerability, and
develop and validate an artificial-intelligence based decision-making tool for determining the locations of food distribution hubs and their food allocation.

Our team is focused on turning research into action. The decision-making tools produced will be disseminated to other disaster-vulnerable food banks throughout the United States, supporting transferable applications of AI-based techniques for food distribution optimization.

This project is part of the CIVIC Innovation Challenge which is a collaboration of NSF, Department of Energy Vehicle Technology Office, Department of Homeland Security Science and Technology Directorate and Federal Emergency Management Agency.

DeimosBC: A Blockchain-based System for Crowdsensing after Natural Disasters

Wed, 01 Sep 2021 00:00:00 +0000

Energy and Emission Prediction for Mixed-Vehicle Transit Fleets Using Multi-Task and Inductive Transfer Learning

Wed, 01 Sep 2021 00:00:00 +0000

Equitable Electric Mobility: Smart Charging and Smart Parking

Sun, 01 Aug 2021 00:00:00 +0000

This Houston-based project establishes a collaborative process with community and commercial technology partners to accelerate the equitable development of accessible fast charging infrastructure and electric vehicle (EV) ownership for low income families by leveraging regional markets of early EV adopters. The novelty of the project lies in a community-driven participatory approach that integrates both social and technical dimensions, bringing transformational change to EV ownership and electrification of smart public transportation. This will be achieved using a data-driven model that integrates real-time data from micro-transit, fixed-route transit, and carpooling services to predict and overcome the uncertainties of traffic conditions, which would result in uncertain travel times and poor reliability. An engaging process establishes collaborative relationships, and identifies adopter indicators and incentives for equitable EV ownership and transit ridership. The project will result in a templet and framework for implementing smart transit hubs, which can be reapplied as best practices at other facilities of the Metropolitan Transit Authority of Harris County (METRO) or other transit systems.

Greater access to new electric mobility technologies requires reduction in economic barriers to EV ownership and transportation services. Communities need to proactively plan for transportation technologies to improve the lives of low-income communities while lowering emissions, increasing transit use, and reducing congestion. The intellectual merits include application of emerging data analytics, smart technologies, and machine learning to tackle a national challenge of equitable access to EV ownership and electric mobility. The project builds a model using artificial neural networks to capture complex, non-linear phenomena and parameterize data for real-time decision making. The broader project impacts include developing a digital ecosystem and incentive programs reflecting the national need to stimulate equitable access to EV ownership; regional policy impact demonstrating collaborative partnerships between cities, EV companies, fast charging developers, and ridesharing services; and research that opens up opportunities to use the data for EV impacts on the energy grid and future smart grid research. This project will identify how the deployment of smart technologies can improve the reliability, comfort, and affordability of transit services.

Transit-Gym: A Simulation and Evaluation Engine for Analysis of Bus Transit Systems

Sun, 01 Aug 2021 00:00:00 +0000

Efficient Data Management for Intelligent Urban Mobility Systems

Sat, 01 May 2021 00:00:00 +0000

A Review and Outlook on Energy Consumption Estimation Models for Electric Vehicles

Mon, 01 Mar 2021 00:00:00 +0000

Selfish Mining Attacks Exacerbated by Elastic Hash Supply

Mon, 01 Mar 2021 00:00:00 +0000

Acceptance Rate: 25%

Minimizing Energy Use of Mixed-Fleet Public Transit for Fixed-Route Service

Mon, 01 Feb 2021 00:00:00 +0000

Acceptance Rate: 21%

Data-Driven Decision Support for Optimizing Cyber Forensic Investigations

Fri, 01 Jan 2021 00:00:00 +0000

Impact Factor: 6.3

Safe and Private Forward-Trading Platform for Transactive Microgrids

Fri, 01 Jan 2021 00:00:00 +0000

The Role of Blockchains in Multi-Stakeholder Transactive Energy Systems

Tue, 01 Dec 2020 00:00:00 +0000

Adversarial Deep Reinforcement Learning based Adaptive Moving Target Defense

Thu, 01 Oct 2020 00:00:00 +0000

AI-Engine for Optimizing Mixed-Fleet Transit Operations

Thu, 01 Oct 2020 00:00:00 +0000

In every public transit system, a trade-off has to be made between concentrating service into very useful routes that serve large numbers of people and spreading service out to ensure that people everywhere have access to at least some service. Improving the efficiency of an existing system while enhancing service in terms of both usefulness and coverage presents considerable challenges. These challenges to operational efficiency are exacerbated by the requirement to provide complementary paratransit services, which are typically characterized by very low efficiency (energy per passenger per mile) and attendant high cost of operation. Our vision is to address these challenges by combining the complementary advantages of fixed- and dynamic-route transit services and seamlessly integrating them. We focus on the following objectives: minimizing energy used per passenger per mile, minimizing passenger wait and trip times, maximizing service coverage, and maximizing the percentage of daily trips serviced by transit. To explore this complex decision space, we will design, implement, and evaluate an artificial intelligence engine, which will enable agencies with mixed-vehicle fleets (EVs, ICEVs, etc.) to operate integrated fixed-dynamic transit services that maximize energy efficiency and make transit more accessible.

This project will bring together a team consisting of people with complementary background in transit operations, transit optimization, simulations, and artificial intelligence. This team will address the above challenges using state-of-the-art artificial intelligence, machine learning, and data-driven optimization techniques. Deep reinforcement learning and Monte-Carlo tree search will form the core of our real-time operational optimization, which will be supported by data-driven optimization for offline planning and by machine learning techniques for predicting transit demand, maintenance requirements, and traffic conditions.

The success of transit optimization and increased ridership depend on the acceptance of the public. Therefore, we are going to develop a novel methodology based on social relational networks to maximize our community outreach. We will augment these methods by providing affordable and complete transit subscription packages and a guidance app leveraged by CARTA’s unique role in managing transit, parking, bike share, and EV car share to support integrated subscription-based access to multiple modes of transportation. To mitigate the risk of operationalization, we will leverage Siemens’ Railigent platform, which provides the capability to manage large-scale geo-spatial data in the cloud and to generate comparative analytics. This partnership also ensures that the software developed here will be of a high technical readiness level and will be provided with required maintenance support to other agencies that face similar problems.

The impact of this project will be immense. The project provides mechanisms to improve transit accessibility with a projected increase from 41% to 73% of residents within ¼ mile of frequent transit service which will reduce personal-vehicle trips taken by those residents, significantly decreasing the energy used for transportation at the city level. The project will also reduce the per passenger per mile energy use and operating costs of the public transit service through operational optimization.

A Privacy-preserving Mobile and Fog Computing Framework to Trace and Prevent COVID-19 Community Transmission