Interested in joining the RISING Lab? Prospective students and postdoctoral researchers should first read the Note to Prospective Students and Post-docs on the lab page.
Three Research Vectors
Our research is organized along three complementary vectors:
- Vector 1: Making IoT systems trustworthy
- Vector 2: Using IoT systems for societal impact
- Vector 3: Building awareness of capabilities and limitations
Portfolio At Glance
Vector 1: Trustworthy IoT Systems
- Secure Microelectronics and Hardware Supply Chains
- Post-Silicon Validation and Debug
- Firmware and Hardware-Software Co-Validation
- Trustworthy Artificial Intelligence and AI Hardware
- Secure Autonomous and Networked Systems
- Formal Verification and Correct-by-Construction Systems
Vector 2: IoT for Societal Impact
- Digital Health, Human Movement, and Wearable Intelligence
- Digital Twins and Intelligent Sensing for Communities and Infrastructure
Vector 3: Awareness and Responsible Adoption
- Exploration Platforms for Technology Awareness
- Undergraduate Research and Research Pathways
- Workforce Development for Trustworthy Computing
Vector 1: Making IoT Systems Trustworthy
IoT systems are increasingly observing the physical world, processing sensitive information, making autonomous decisions, and controlling safety-critical processes. Their trustworthiness cannot be established by protecting only one component. A secure chip can still execute vulnerable firmware; a reliable sensor can feed information to a compromised AI model; a well-designed autonomous controller can be misled by malicious communication.
Our research addresses trustworthiness across the complete system stack: from semiconductor devices and hardware architectures to firmware, artificial intelligence, networks, and autonomous behavior.
1. Secure Microelectronics and Hardware Supply Chains
The challenge
Modern electronic systems are assembled from processors, memories, accelerators, communication components, and other intellectual-property blocks obtained through a global supply chain. Chips may pass through multiple organizations during design, fabrication, testing, packaging, and deployment. This distributed ecosystem enables innovation, but it also creates opportunities for intellectual-property theft, unauthorized modification, reverse engineering, malicious components, counterfeit devices, and exposure of sensitive information.
Emerging chiplet-based and three-dimensional integrated systems make this challenge more urgent. A single product may combine components developed by different vendors, manufactured at different facilities, and connected through complex internal communication fabrics. Security mechanisms must protect the resulting system without making small and resource-constrained devices prohibitively expensive.
Our accomplishments
The RISING Lab has developed a sustained body of research on architectures for specifying, implementing, validating, and managing security policies in System-on-Chip designs. Our earlier work established centralized and configurable approaches for enforcing policies such as access control and information-flow protection, including in systems containing untrusted hardware components.
More recent research has produced security architectures and methodologies that protect chips throughout their supply chains. These include SENTRY, an infrastructure for protecting System-on-Chip designs against supply-chain attacks; MCSE, a methodology for designing configurable and customizable security engines for resource-constrained devices; secure provisioning methods for protecting sensitive assets in untrusted testing facilities; and TREEHOUSE, an asset-management infrastructure for three-dimensional integrated circuits. We have also developed tools ObNoCs, PoTeNt, and related techniques that obfuscate or reconfigure on-chip communication networks to make hardware designs more difficult to reverse engineer.
Our work has additionally contributed to systematic evaluation of logic locking, redaction, state-space transformation, interconnect protection, and other hardware-security techniques. These projects have brought together expertise in hardware architecture, formal methods, semiconductor design, fabrication, testing, and system integration.
Our vision
Our long-term goal is to make security a configurable and evolvable property of the chip rather than a collection of fixed protections added late in the design process. We envision microelectronic systems that can express their security requirements, monitor whether those requirements remain satisfied, and adapt their protections as threats, workloads, and deployment conditions change.
- Security of heterogeneous chiplets and three-dimensional integrated systems
- Protection of on-chip interconnects and communication fabrics
- Configurable security for low-cost and resource-constrained devices
- Secure provisioning, testing, deployment, and field update
- Detection of malicious or compromised hardware components
- AI-assisted generation and validation of hardware security policies
- Security architectures that remain effective across the complete system life cycle
This research creates opportunities for collaboration among semiconductor companies, system integrators, government agencies, and researchers in hardware architecture, manufacturing, packaging, testing, and cybersecurity.
2. Post-Silicon Validation and Debug
The challenge
Modern semiconductor chips contain billions of transistors and integrate processors, memories, accelerators, communication fabrics, and specialized hardware components. Despite extensive verification before fabrication, some defects become visible only after a chip has been manufactured and begins executing realistic workloads.
Finding such defects is especially difficult because engineers can observe only a small fraction of the chip’s internal behavior. Internal storage and communication bandwidth are limited, failures may be intermittent, and the evidence collected during an execution may be far removed from the underlying cause. Engineers must therefore decide which internal signals to observe, when to capture them, and how to reconstruct the sequence of events that produced a failure.
Our accomplishments
Post-silicon validation has been a distinctive and sustained component of the RISING Lab’s research. We have developed methods for selecting informative internal signals, improving the observability of fabricated chips, analyzing execution traces, and localizing failures from limited post-silicon evidence. Our research has also explored how information developed during pre-silicon verification can be reused after fabrication rather than treating the two stages as disconnected activities.
A central contribution of this work is the development of systematic techniques for deciding what information should be collected from a running chip. Because only a small subset of internal signals can be recorded, effective trace selection can substantially determine whether a failure can later be understood. We have investigated both architectural support and analysis methods for making this limited information more useful for validation and debug.
The novelty and impact of this research have been recognized through best-paper nominations at both the Design Automation Conference (DAC) and the International Conference on Computer-Aided Design (ICCAD). The work provides an important foundation for addressing validation challenges that emerge only in fabricated hardware and under realistic system execution.
Our vision
We envision post-silicon validation as a principled and increasingly automated activity rather than a predominantly manual search through incomplete traces. Future validation platforms should identify the most informative internal observations, relate observed symptoms to likely causes, and adapt their data collection as engineers learn more about a failure.
- Post-silicon observability and trace-signal selection
- Efficient on-chip collection and compression of execution information
- Failure detection and localization from incomplete traces
- Reconstruction of long and complex execution histories
- Reuse of pre-silicon models, properties, and verification results after fabrication
- Integration of formal reasoning, data analysis, and machine learning for debug
- Validation of heterogeneous Systems-on-Chip and accelerator-based architectures
- Architectural support for field diagnosis and lifecycle assurance
- Methods for distinguishing design defects from manufacturing, environmental, and aging-related failures
This research is especially relevant to semiconductor companies, processor and accelerator developers, electronic-design-automation providers, government laboratories, and organizations responsible for validating advanced microelectronic systems.
3. Firmware and Hardware-Software Co-Validation
The challenge
Firmware is the low-level software that initializes, configures, and controls electronic hardware. It manages functions such as startup, power, communication, security, sensing, and device operation. Unlike conventional application software, firmware interacts directly with processors, memories, peripherals, interrupts, resets, clocks, and specialized hardware components.
Many system failures cannot be attributed to hardware or software in isolation. They arise from an incorrect assumption made by one component about the other, a rare ordering of asynchronous events, an unexpected device state, or a mismatch between the firmware and an evolving hardware specification. Conventional software testing often abstracts away the hardware details that trigger these failures, while hardware verification may simplify or omit the firmware that controls the system.
Hardware-software co-validation therefore requires both components to be analyzed together within a sufficiently realistic execution environment.
Our accomplishments
The RISING Lab has developed methods that treat hardware and firmware as a coupled system rather than validating them independently. FirVer applies concolic testing – a combination of concrete execution and symbolic reasoning – to systematically explore firmware binaries and uncover behaviors that are difficult to reach through conventional testing. CASTLE provides an architectural approach for protecting the integrity of System-on-Chip firmware.
Our formal-analysis research examines interactions among firmware, processors, peripherals, interrupts, and other hardware components. This work provides methods for reasoning about control flow, information flow, asynchronous events, and security-sensitive behavior across the hardware-software boundary.
We have also developed VirSoC, a methodology for automatically synthesizing virtual System-on-Chip environments. Such platforms allow firmware to execute against models of the hardware before the final chip is available. They make it possible to explore rare interactions, construct repeatable experiments, and validate firmware without requiring continuous access to physical prototypes.
Together, these efforts establish complementary capabilities in firmware exploration, formal reasoning, security architecture, and virtual prototyping. Our current co-validation research focuses on pre-silicon and virtual-platform environments; it is distinct from the lab’s post-silicon validation program.
Our vision
We want to establish a systematic science of hardware-software co-validation in which firmware is analyzed within the hardware environment that gives its operations meaning. Rather than validating hardware and software independently and integrating them late in development, we envision continuous co-validation beginning with early architectural models and continuing as both the hardware and firmware evolve.
An important aspect of this vision is support for changing specifications. Hardware defects are sometimes mitigated through firmware modifications, and firmware requirements may evolve as hardware behavior becomes better understood. Validation methods must therefore distinguish implementation errors from legitimate changes and determine whether the evolving hardware-software system continues to satisfy its intended requirements.
- Validation of binary-only and third-party firmware
- Formal semantics for hardware-firmware interaction
- Systematic exploration of interrupts, resets, timing, concurrency, and power transitions
- Analysis of dynamically loaded and highly optimized low-level firmware
- Automated synthesis of virtual System-on-Chip environments
- Integration of formal verification, concolic execution, fuzzing, and simulation
- Validation across multiple processors, peripherals, and privilege domains
- Runtime monitoring of firmware integrity and control flow
- Co-validation when hardware behavior or firmware specifications evolve
- Assurance of firmware used in automotive, communication, medical, security, and AI systems
This research is particularly relevant to organizations developing complex Systems-on-Chip, embedded devices, automotive electronics, communication platforms, medical devices, and other systems in which hardware and firmware must operate correctly together.
4. Trustworthy Artificial Intelligence and AI Hardware
The challenge
Artificial intelligence is becoming part of the control and decision-making infrastructure of IoT systems. AI models interpret sensor data, detect anomalies, manage resources, control vehicles, and support health decisions. However, these models can be manipulated through adversarial inputs, corrupted computations, malicious model changes, or attacks on the hardware on which they execute.
Specialized AI accelerators introduce additional concerns. Multiple applications may share an accelerator, physical faults may corrupt intermediate computations, and power or timing behavior may reveal sensitive information. Securing only the neural-network model is therefore insufficient; trustworthiness must extend from the model and its training assumptions down to the underlying architecture and physical implementation.
Our accomplishments
Our research has examined integrity and evasion attacks on deep-learning systems and developed methods for detecting and responding to them. Geralt supports real-time detection of evasion attacks, while subsequent work combines enhanced attack detection with model updating so that the system can adapt after previously unseen attacks are discovered.
We developed AroMa to evaluate stealthy integrity attacks on deep-learning applications executing on shared, multi-tenant accelerators. We have also studied targeted fault-induced misclassification, the security implications of integrating AI accelerators into microelectronic systems, and limitations in current approaches for evaluating neural networks under integrity attacks.
Complementing this security work, we have developed correct-by-construction methods for custom accelerator microarchitectures and formalized accelerator behavior for neural-network scheduling and optimization. Recent research also explores the use of large language models and policy-governed AI agents to generate hardware security properties and assist with microelectronic assurance.
These projects connect expertise in machine learning, computer architecture, physical attacks, formal methods, and embedded systems, and have often required close collaboration among researchers working at different layers of the AI computing stack.
Our vision
Our objective is to create AI systems whose trustworthiness can be evaluated across the complete computing stack. Such systems should be able to recognize when they are operating outside their assumptions, detect corrupted computations, recover from attacks or faults, and provide meaningful evidence about the reliability of their decisions.
- Fault and side-channel attacks on AI accelerators
- Security of shared and multi-tenant AI platforms
- Runtime detection, diagnosis, and recovery from corrupted AI behavior
- Resilient and correct-by-construction accelerator architectures
- Assurance of large language models and autonomous AI agents
- AI-assisted hardware and firmware security analysis
- Evaluation methodologies that connect physical faults to application-level consequences
- Techniques for distinguishing normal model uncertainty from malicious manipulation
This program provides a natural collaboration point for researchers and sponsors working in trustworthy AI, edge computing, semiconductor security, autonomous systems, and safety-critical machine learning.
5. Secure Autonomous and Networked Systems
The challenge
Autonomous systems depend on information collected from sensors and received from other devices. A connected vehicle, for example, may combine cameras, ranging sensors, wheel-speed sensors, wireless messages, maps, and AI-based perception to make decisions about braking, steering, lane changes, and vehicle spacing.
An attacker does not necessarily need to compromise the vehicle’s internal computer. Manipulating a sensor or transmitting a misleading message may be sufficient to produce unsafe physical behavior. The challenge is therefore not only to prevent attacks but also to enable the system to continue operating safely when some of its information sources become unreliable.
Our accomplishments
The RISING Lab has developed a broad portfolio of attacks, defenses, evaluation methods, and experimental platforms for connected and autonomous vehicles. Our work on RACCON, REDEM, and RECAP developed machine-learning-based approaches for detecting and mitigating communication and multi-channel perception attacks in cooperative adaptive cruise-control systems. CAVeliEr supports automated security evaluation of connected-vehicle applications, while DRiFt addresses resilient coordination of distributed vehicle fleets under communication attacks.
We have investigated attacks on ranging sensors, wheel-speed sensors, ride-height sensors, steering systems, vehicle-to-infrastructure communication, 5G-enabled cooperative control, lane-changing decisions, and model-predictive vehicle platoons. We have also developed physical, virtual, and mixed-reality platforms that allow these attacks and defenses to be examined safely and systematically.
Our work extends beyond security to automotive functional safety, vehicle-health monitoring, and accessible autonomous transportation for passengers with visual disabilities. These projects have relied on interdisciplinary collaboration spanning cybersecurity, transportation, communication, control, automotive engineering, artificial intelligence, accessibility, and human-centered design.
Our vision
We envision autonomous and multi-agent systems that remain safe even when some sensors, communication channels, or collaborating agents are faulty, compromised, or uncertain. Achieving this requires moving beyond isolated intrusion detection toward system-level resilience: the ability to understand the operational consequence of suspect information and adapt before it causes physical harm.
- Resilient perception and control under simultaneous sensor and communication attacks
- Security of cooperating fleets of vehicles, drones, and robots
- Assurance of multi-agent and human-machine teams
- Integration of cybersecurity with functional safety
- Validation of autonomous behavior under rare and adversarial conditions
- Security of 5G, V2X, and future vehicle communication
- Runtime estimation of whether an autonomous decision can be trusted
- Digital-twin and mixed-reality environments for large-scale evaluation
Potential collaborators include automotive manufacturers, transportation agencies, communication providers, autonomy researchers, defense organizations, and developers of robotic and cyber-physical systems.
6. Formal Verification and Correct-by-Construction Systems
The challenge
Testing can demonstrate that a system works in the situations that have been examined, but it generally cannot explore every possible execution. Rare combinations of events may remain undiscovered until deployment, where they can lead to costly failures or exploitable vulnerabilities.
Formal verification addresses this limitation by representing systems and requirements mathematically and using machine-checked reasoning to establish whether the requirements hold. The principal challenge is scalability: realistic hardware and software systems are far too complex for any single verification technique.
Our accomplishments
Formal methods form the intellectual foundation of the RISING Lab. Our work has advanced theorem proving, model checking, predicate abstraction, compositional reasoning, machine-code verification, microarchitecture assurance, and combinations of deductive and algorithmic techniques.
We developed scalable frameworks for certifying behavioral and high-level synthesis, including methods for proving that aggressively optimized hardware implementations remain equivalent to their high-level specifications. We have applied formal reasoning to concurrent and reactive systems, industrial memory designs, processor and System-on-Chip architectures, security policies, firmware interactions, communication protocols, and accelerator microarchitectures.
This research has produced a book on scalable formal verification, books on System-on-Chip and hardware security, and a sustained series of methods for integrating general mathematical reasoning with automated decision procedures. It has also involved longstanding collaboration with researchers in formal methods, programming languages, hardware design, security, and industrial verification.
Our vision
Our vision is not to apply formal verification only after a system has been designed. Instead, we are creating correct-by-construction systems in which specifications, assumptions, verification evidence, and runtime monitoring evolve together with the design.
- Compositional verification of hardware, firmware, software, and networks
- Integration of theorem proving, model checking, testing, and learning
- Verification of dynamically changing and adaptive systems
- Formal foundations for trustworthy autonomous and AI systems
- Automatic generation of properties, abstractions, and proof guidance
- Translation of design-time guarantees into runtime monitors
- Human-understandable explanations of formal assurance results
- Architectures designed from the outset to support analysis and verification
This program connects the lab’s foundational expertise to every other component of the research portfolio.
Vector 2: Using IoT Systems for Societal Impact
Our second research vector asks how connected sensing, intelligent computation, and digital representations of physical systems can address consequential societal problems.
The goal is not simply to apply an existing IoT platform to a new domain. We collaborate with domain experts to understand what can be measured, what decisions the measurements can support, and what technical limitations could make the conclusions misleading. Trustworthiness therefore remains a central requirement even when societal impact is the primary motivation.
7. Digital Health, Human Movement, and Wearable Intelligence
The challenge
Most health and movement assessments are collected during occasional clinical visits, even though disease progression, rehabilitation, and changes in mobility occur continuously. Wearable devices and connected sensors provide an opportunity to observe physiological, behavioral, lifestyle, gait, and movement indicators over longer periods. However, transforming these data into medically meaningful information requires careful modeling, personalization, and validation.
Our accomplishments
The RISING Lab has developed wearable, machine-learning, language-based, and digital-twin approaches for health monitoring. MASC investigates wearable detection of infectious disease, while related work explores continuous monitoring of physiological parameters and wearable detection of COVID-19 in symptomatic patients. ScreeningPaL uses caregiver-provided free-text descriptions to support early screening for autism.
Our liver-health research combines longitudinal lifestyle data, biomarker forecasting, machine learning, and digital-twin simulation. Twin-SCan estimates liver biomarkers through machine learning and digital-twin analysis, and our work on multivariate time-series forecasting investigates noninvasive detection of liver disease from longitudinal data.
Human gait and mobility form another active and growing research direction. We have developed a machine-learning approach for real-time gait analysis and a wearable system for knee-rehabilitation monitoring that provides immediate feedback. The gait-analysis work received a Best Paper Runner-Up Award at IEEE SmartIoT 2025. Together, these projects explore how wearable sensing and intelligent analysis can move gait assessment beyond brief clinical observations toward continuous, objective, and personalized measurement in rehabilitation and everyday settings.
Collectively, this research has contributed to publications across engineering, sensing, biomedical informatics, and medical venues, as well as books on personalized health monitoring and digital twins. The projects have been carried out through collaboration among computing researchers, clinicians, biomedical engineers, rehabilitation specialists, and other health-domain experts. Such partnerships are essential for ensuring that the technical questions, data, evaluation criteria, and intended uses reflect meaningful healthcare needs.
Our vision
We envision personalized health and human-movement systems that combine wearable measurements, longitudinal records, physiological and biomechanical knowledge, machine learning, and digital twins to provide timely and understandable information. These tools are not intended to replace healthcare professionals. Rather, they can help identify meaningful changes earlier, support continuous rehabilitation and mobility assessment, and provide clinicians with information that is difficult to obtain from isolated measurements.
- Personalized digital twins for disease progression and intervention
- Continuous and noninvasive biomarker estimation
- Wearable and machine-learning systems for gait, mobility, and rehabilitation assessment with real-time feedback
- Personalized and longitudinal modeling of gait and human mobility
- Integration of physiological, behavioral, and lifestyle data
- Trustworthy health-focused machine learning and language models
- Privacy and security of Internet-of-Medical-Things systems
- Methods for communicating uncertainty to patients and clinicians
- Deployment with clinical and community partners
We welcome collaboration with healthcare researchers, clinicians, rehabilitation specialists, biomedical engineers, public-health organizations, and developers of wearable technologies.
8. Digital Twins and Intelligent Sensing for Communities and Infrastructure
The challenge
Many societal problems are difficult to study directly. Contamination may be detected only after exposure has occurred; industrial failures may be too expensive or hazardous to reproduce; and changes in energy or transportation infrastructure may be impractical to test physically.
Digital twins provide virtual representations of physical devices, environments, and processes that evolve using data from their real-world counterparts. Combined with intelligent sensing, they can help detect problems, explore alternatives, and predict future behavior without disrupting the physical system.
Our accomplishments
The RISING Lab has developed digital-twin technologies for distributed IoT applications, automotive systems, healthcare, energy, manufacturing, and environmental monitoring. ViSE provides a digital-twin environment for exploring automotive functional safety and cybersecurity, while DT-IoMT defines a reference model for secure digital twins in the Internet of Medical Things.
AQuaTwin applies the digital-twin concept to early detection of water contamination. Our work has also examined digital-twin tools for smart manufacturing and Industry 4.0, IoT-driven energy systems, healthcare delivery, and vehicular prototyping. These efforts have contributed to a book on digital twins for distributed IoT applications and to surveys and frameworks that help organize this rapidly growing field.
Our intelligent-sensing research additionally includes nuclear quadrupole resonance for substance detection and AI-enabled approaches to mosquito baiting and monitoring. We have explored IoT systems for vehicle-health monitoring, food-donation distribution, and accessible transportation, illustrating the breadth of problems that can be addressed when sensing and computation are designed around a specific societal need.
This work is inherently interdisciplinary. We have collaborated with researchers and practitioners in environmental science, manufacturing, healthcare, transportation, energy, sensing, and public health to connect technical innovation with realistic requirements and meaningful evaluation.
Our vision
Our long-term objective is to create reusable but trustworthy digital-twin and sensing technologies that can be adapted to new societal challenges without rebuilding the complete infrastructure for every application.
- Digital twins for environmental and public-health monitoring
- Smart manufacturing and resilient industrial infrastructure
- Energy-system monitoring, prediction, and optimization
- Detection of water, chemical, and biological hazards
- Sustainable transportation and accessible mobility
- Integration of heterogeneous and imperfect sensor data
- Secure synchronization between physical systems and their digital twins
- Quantification of confidence, fidelity, and uncertainty in twin predictions
- Community-centered deployments developed with domain and public partners
These projects provide opportunities for interdisciplinary collaboration with experts in health, environmental science, manufacturing, agriculture, energy, transportation, sensing, and public policy.
Vector 3: Building Awareness of Capabilities and Limitations
The third research vector addresses how people understand, evaluate, and engage with increasingly complex connected and intelligent systems. These technologies often appear capable and authoritative even when they depend on incomplete information, simplifying assumptions, or operating conditions that may not hold in practice.
Our research develops exploration platforms, educational models, and workforce pathways that help different communities understand both the potential and the limitations of emerging technologies. The objective is not simply to disseminate technical knowledge, but to enable informed experimentation, responsible use, and broader participation in the development of trustworthy systems.
9. Exploration Platforms for Technology Awareness
The challenge
Complex IoT, automotive, and AI systems are difficult to understand through descriptions alone. Their most important vulnerabilities often arise from interactions among sensors, software, communication, machine learning, and physical behavior. Physical systems may also be too expensive, inaccessible, or dangerous for widespread experimentation.
Effective awareness therefore requires more than simplified demonstrations. Users need environments in which they can explore realistic system behavior, vary assumptions, observe failures, and connect an apparently abstract cyber event to its physical consequences.
Our accomplishments
The RISING Lab has developed a family of physical, virtual, and mixed-reality platforms for exploring automotive, IoT, and AI safety and security. These environments allow users to manipulate sensors, communication channels, intelligent models, and system components and then observe how those changes affect a vehicle or another physical system.
Our platforms support exploration of ranging-sensor attacks, wheel-speed and MEMS sensor manipulation, vehicular communication attacks, machine-learning vulnerabilities, functional-safety failures, and the relationship between cyber events and physical behavior. They make experiments possible without requiring every learner or institution to own a costly test vehicle or specialized laboratory.
This work is itself a research program in cybersecurity education and technology awareness. We study how immersive and interactive experiences complement conventional instruction, how complex vulnerabilities can be represented without losing technical meaning, and how exploration platforms can reach students, engineers, decision-makers, and the broader community.
Our vision
We envision scalable exploration environments that make emerging technologies open to investigation by diverse communities. The same platform should be able to support education, public engagement, reproducible research, engineering evaluation, and informed discussion of the assumptions and limitations of intelligent systems.
- Immersive exploration of hardware, AI, automotive, and IoT security
- Digital twins and mixed-reality environments for safety and security awareness
- Platforms that connect cyber events with visible physical consequences
- Remote and reproducible access to specialized experiments
- Research on how interactive exploration changes understanding and decision-making
- Tools that help the broader community evaluate both capabilities and limitations
10. Undergraduate Research and Research Pathways
The challenge
Early exposure to authentic research can strongly influence whether undergraduate students pursue advanced study and careers in emerging technical fields. Yet many students, particularly those at institutions with limited research infrastructure, have few opportunities to participate in open-ended research or to see how their technical training can address consequential societal problems.
Our accomplishments
Our Research Experiences for Undergraduates (REU) program has been introducig students from institutions across the country to research in secure, accessible, and sustainable transportation. Participants investigate problems arising from autonomy, electronics, sensing, communication, and software in modern transportation systems.
Students work on mentored projects involving the safety and security of autonomous vehicles, resilient connected transportation, accessible mobility, intelligent sensing, and related cyber-physical systems. They also receive preparation in research methods, experimentation, technical communication, graduate education, and professional development.
The program is designed to produce more than a collection of summer projects. It helps students learn how research questions are formed, how evidence is developed, how results are communicated, and how interdisciplinary research can serve societal needs. By recruiting nationally, it broadens access to experiences that may not be available at a student’s home institution.
Our vision
We seek to build sustained pathways that connect undergraduate discovery, graduate study, and careers in trustworthy autonomous and cyber-physical systems. Our broader goal is to expand who participates in research while preparing students to work across the technical and societal boundaries that shape modern transportation.
- National access to mentored research in trustworthy transportation
- Research experiences spanning safety, security, accessibility, and sustainability
- Pathways from undergraduate research to graduate education
- Supportive research communities for students from varied institutions and backgrounds
- Integration of technical research, communication, and professional development
11. Workforce Development for Trustworthy Computing
The challenge
The workforce needed to develop secure hardware, trustworthy AI, and resilient IoT systems must include people entering through many different pathways. Veterans transitioning into civilian careers, working professionals developing new expertise, and educators preparing future technicians and engineers bring different strengths and face different barriers. A single educational model cannot serve all of these communities effectively.
Our accomplishments
Through VETS-HASTE, the RISING Lab provides veterans with focused, hands-on preparation for careers in hardware security and the semiconductor ecosystem. The program combines practical technical training, mentoring, engagement with industry, and experiential learning using realistic hardware-security tools, attacks, and defenses. It builds on veterans’ operational experience while providing a pathway into a specialized and rapidly growing civilian field.
We are also developing educator-focused approaches that prepare community-college faculty to introduce artificial intelligence, AI safety, and the limitations of intelligent systems within their courses. This train-the-educator model creates a multiplier effect: participating faculty can carry the concepts, experiments, and curricular materials to students across multiple disciplines, institutions, and academic years.
Together, these efforts investigate how audience-specific program design, authentic technical practice, mentoring, and hands-on learning can create effective pathways into trustworthy computing. Their impact lies not only in increasing the number of trained individuals, but also in broadening who can participate in the development and responsible deployment of advanced technologies.
Our vision
We envision adaptable workforce-development models that connect education with realistic engineering practice, research opportunities, internships, and long-term careers. By recognizing the distinct needs of different learner communities, we seek to expand the technical workforce while keeping security, safety, and responsible technology use central to its preparation.
- Experiential hardware-security pathways for veterans and career-transition learners
- Faculty development that scales AI-safety education through community colleges
- Hands-on curricula grounded in realistic tools, attacks, and defenses
- Connections among training, mentoring, internships, research, and employment
- Evaluation of workforce models for learners with different backgrounds and goals
Connecting the Three Vectors
Trustworthiness provides the foundation. We develop secure architectures, validation methods, formal reasoning, resilient AI, and system-level defenses.
Societal applications provide the purpose. We use these foundations to address challenges in health, transportation, environmental monitoring, manufacturing, energy, accessibility, and infrastructure.
Awareness enables responsible adoption. We create exploration platforms, undergraduate research pathways, and targeted workforce programs that help people understand both the potential and the limitations of these technologies.
Many projects naturally cross these boundaries. A digital twin for healthcare contributes to societal impact, but it also raises questions about security, privacy, fidelity, and uncertainty. An automotive attack platform supports education, but it also enables research on resilient control and autonomous-system assurance. An AI-assisted hardware-security tool advances trustworthiness while changing how future engineers interact with complex systems.
Collaboration also connects the three vectors. Foundational researchers help establish assurance principles; industry and government partners contribute realistic systems and operational requirements; and domain experts ensure that societal applications address consequential problems and are evaluated appropriately. This interconnected model allows the RISING Lab to pursue foundational research while remaining engaged with practical systems and real-world needs.
Opportunities to Engage with the RISING Lab
The RISING Lab has substantial experience participating in and leading interdisciplinary and multi-institutional research efforts involving universities, industry partners, government organizations, and application-domain experts. We welcome new collaborations that combine our expertise in trustworthy computing with complementary capabilities, realistic systems, specialized data, experimental facilities, or deployment opportunities.
Potential partners include academic researchers, semiconductor and technology companies, automotive and transportation organizations, healthcare and environmental researchers, government agencies, national laboratories, and community organizations.
Prospective students can contribute from many disciplinary backgrounds, including computer engineering, electrical engineering, computer science, cybersecurity, formal methods, artificial intelligence, embedded systems, sensing, biomedical engineering, transportation, and human-centered computing. Those interested in joining the lab should review the Note to Prospective Students and Post-docs before contacting us.
We also warmly welcome engagement with the lab through foundational research, joint proposals, sponsored projects, prototype development, technology evaluation, student internships, shared experimental platforms, and transition of research methods into practical systems.
|
Learn more: People | Publications and collaborations | Note to Prospective Students and Post-docs |
Detailed examples of our research projects, publications, coauthors, and collaborative activities are available on the RISING Lab Publications page.
Across all these activities, our guiding objective remains the same: to find ways to develop connected and intelligent systems that people can trust, use such systems to improve society, and understand well enough to deploy responsibly.