top of page


Our research is focused on intelligent devices that physically assist or monitor human motion.


SAFE Orthosis


Neuromuscular disorders such as stroke often cause walking impairments. Conventional exercise-based gait rehabilitation protocols involve multiple training sessions supervised by one or more physical therapists. This process is physically demanding for therapists and expensive for the healthcare system. Rehabilitation robots have the potential to automate these exercise protocols.

Compared to hip and knee, the ankle joint contributes the largest propulsive moment during walking. For this reason, many prototypes of powered ankle-foot orthoses (AFO) have been developed in the last decade for use in gait rehabilitation. Yet, most of these designs are complex, heavy, and tethered to an actuation station.  The Stevens Ankle-Foot Electromechanical (SAFE) orthosis was developed with the goal of providing in-home, self-administered ankle rehabilitation. This device is built off of a modified articulated AFO and adds only 1.04 kg of extra weight to the patient’s lower leg.  The uniqueness of the SAFE orthosis relies on the possibility to quickly swap between two actuation approaches: fully-active mode, which employs two antagonistic motors to provide plantar- and dorsiflexion torques, and semi-active mode, which achieves the same goal by means of a single motor and an antagonistic spring. This feature allows us to characterize the device’s performance (e.g., closed-loop bandwidth) under the two activation modes in bench tests as well as human experiments, with the goal of paving the way for a new generation of simpler, more affordable, yet fully-functional powered AFOs.  In collaboration with Dr. Karen Nolan at Kessler Foundation’s Human Performance Movement & Rehabilitation Engineering, we are about to start pilot tests on chronic stroke survivors.



Quantitative gait analysis is a powerful tool for physicians treating patients with gait disorders.  Athletic trainers often rely on assessments of the running gait when coaching professional athletes who are recovering from an injury or want to improve their performance.

Quantitative gait analysis typically requires specialized laboratory equipment such as optical motion capture systems and treadmills instrumented with force plates. For this reason, the use of gait analysis is current hampered by high operating costs and lack of portability.


In recent years, instrumented footwear has been developed for portable gait assessments. Compared to traditional laboratory equipment, these new systems are more affordable and versatile. However, the number of parameters they can assess is limited, and their accuracy is usually poor.


By adopting custom calibration algorithms, SportSole can achieve higher levels of accuracy than other footwear-based devices while relying on mid-grade, cost-effective sensors. The system consists of insoles instrumented with inertial and piezo-resistive sensors. It is capable of measuring a rich set of spatiotemporal gait parameters (stride length, foot-ground clearance, foot trajectory, cadence, single and double support times, symmetry ratios and walking speed) as well as kinematic parameters (namely, dynamic plantar maps and center of pressure trajectories) during walking and running tasks.



PediShoe, the first instrumented footwear specifically designed for pediatric populations, was developed with the Columbia University's ROAR Lab. The system consists of instrumented sandals, each one weighting less than 100g, which can accurately measure spatial and temporal gait parameters in toddlers and children.

We are currently validating accuracy and precision of the system with typically developing children and clinical populations. Using this tool, we are interested in assessing  gait and balance of children with movement or developmental disorders (e.g., Cerebral Palsy and Autism Spectrum Disorder) in the unconstrained environment.

In the near future, easy-to-use, portable and affordable devices like PediShoe have the potential to support diagnosis, prognosis, and monitoring of movement disorders in pediatric populations.

Smart Crutches


Crutches enable patients to partially transfer their weight bearing to the upper extremities, thus decreasing the demand on weak or impaired legs. The improper use of these walking aids, however, can lead to secondary injuries and prolong the recovery time. 

Applying too much weight on the lower extremities after surgery can disrupt the operated tissues and bones, while overloading the shoulders can cause suprascapular neuropathy, shoulder arthropathy, and axillary artery occlusion. Due to limited time and limited clinical personnel, training patients on how to properly use crutches as their patterns of recovery progress is often unfeasible. Moreover, it is currently impossible to monitor how patients use crutches in their daily life.


IoT wearable technology offers a viable solution to these needs. The Smart Crutches are a pair of forearm crutches retrofitted with custom-made load cells, inertial sensors, and vibro-tactile transducers. Using the embedded electronics, they can estimate 3D ground reaction forces in real-time and alert the patient when an improper use is detected.

Besides their use as a new training and monitoring device, we envision the application of the Smart Crutches as a research tool to study human-robot physical interaction in portable robotic exoskeletons for gait training (e.g., ReWalk and EksoGT), and to develop new human-in-the-loop control strategies for those robots.

We are currently working on the second design iteration of the system, using force plates and an optical motion capture system to validate its performance.

Personal Flotation Device


Inflatable Personal Flotation Devices (PFDs) are common maritime safety equipment, but current models do not fully meet the needs of vessels' crews and passengers. The two main reasons for this are: discomfort to the user and premature activation in water

At the Stevens' WRS and Davidson Laboratories, we have been working on a new PFD concept, which incorporates multiple inflatable bladders and embedded logic, sensing, and deployment equipment. Inflation of the bladders is controlled in real-time to ensure that unconscious wearers are turned face up in the water. The Smart PDF device is currently undergoing testing in the Davidson Laboratory's High Speed Towing Tank.

Cable-Driven Parallel Robots


Cable-Driven Parallel Robots (CDPR) belong to a special subclass of parallel manipulators in which the moving platform is supported in parallel by cables. Generally, each cable is reeled on an actuated pulley and has the other extremity fixed to an attachment point on the moving platform.

A main difference between CDPRs and common parallel robots is that the supporting cables can only apply tensile forces on the moving platform. For this reason, controlling CDPRs is challenging.

  In the WRS lab, we study analytical tools that allow designers to optimize the robot layout for a desired set of requirements, and develop reconfigurable CDPRs that can adapt their layout online, based on the task. We also explore trajectory planning and control algorithms that allow CDPRs to operate beyond the boundaries of their workspace.

bottom of page