Human-Centered Engineering for Mobility Lab

Understanding how people move to design assistive technologies that work with them.

We combine experimental and computational approaches in biomechanics and motor control to understand how wearable robots affect movement. We strive to develop adaptive systems that respond to the values and lived experiences of the communities we work with.

Research Directions

Locomotor Control Beyond the Lab

Moving through everyday environments requires reconciling competing demands, such as conserving energy while moving quickly or crossing uneven terrain while staying balanced. We are building models that infer those priorities from measured movement^§. We are relating these priorities to the coordination strategies that express them and studying how both change with aging and neurological gait conditions.

wearable sensing · musculoskeletal simulation · inverse optimal control · machine learning

Person descending outdoor stairs with wearable sensors on legs and face, beside a musculoskeletal model ascending stairs and inset plots of movement dynamics and shifting control priorities.

Holistic Exoskeleton Optimization

Mobility aids shape both how people move and how they perceive that movement, with responses changing over time. We are designing lower-limb exoskeleton control strategies that keep adapting as people's behavior changes, personalizing assistance to multiple objectives^§ and considering perceptual outcomes alongside physical performance. Our work examines the tradeoffs among these objectives and what leads devices to be adopted rather than abandoned as people engage with technology over longer timescales.

exoskeleton hardware and control · human-in-the-loop optimization · psychophysics

Person walking outdoors in a hip exoskeleton, with arrows tracing a human-in-the-loop optimization cycle: assistance settings, measured responses, an optimization landscape, and updated torque profiles.

Learning New Movement Strategies

Learning a new motor skill depends on interpreting cues from the body and environment, a process that well-designed feedback can accelerate. We are developing real-time biofeedback paradigms that guide people toward strategies suited to their own movement, rather than a single prescribed target^§. We are investigating which signals meaningfully aid learning under which conditions and how skills learned with feedback can endure without it, supporting independence in clinical and community settings.

real-time biofeedback · haptic interfaces · practice design · motor learning and retention

Person walking on a treadmill with ankle exoskeletons while a monitor provides real-time visual biofeedback about their movement.

Our Broader Research Framework

Supporting mobility with technology requires more than building a device. We need ways to understand people’s priorities in the context of their own bodies and everyday environments, represent those priorities in our design objectives, and evaluate to what extent technology actually supports them. This framework provides a structure for connecting these considerations across the research process. Although shown as stages, the process is iterative, with insights from each stage shaping the others.

Listen

Learning from people's own accounts of their mobility, in the context of their values and goals.

Encode

Representing insights as qualitative and quantitative data that can be systematically analyzed and used to develop testable models.

Integrate

Connecting models of biomechanical, neural, and psychosocial processes to understand how they jointly shape human behavior.

Design

Creating physical systems and interfaces whose form and function are informed by integrated models of behavior.

Operationalize

Developing personalized control strategies that adapt device behavior to each person as they move.

Validate

Evaluating effectiveness beyond the laboratory, in the context of people’s everyday lives.

Each of these stages represents an important area of inquiry and collaboration. While our published and ongoing work focuses primarily on Encode and Operationalize—inferring movement priorities from behavior and building control strategies that respond to them—fully realizing this vision will require expertise and collaboration across the entire framework.