Jennifer K. Leestma
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Jennifer K. Leestma has a total of 25 co-authors. Here are the top co-authors by number of shared publications:
- Aaron J. Young
Georgia Institute of Technology
7 shared publications - Gregory S. Sawicki
School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, GA, USA
7 shared publications - Anirban Mazumdar
Georgia Institute of Technology
2 shared publications - Christoph P. O. Nuesslein
Georgia Institute of Technology
2 shared publications - Courtney R. Smith
Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology, Atlanta, USA
2 shared publications - Felicia Davenport
Georgia Institute of Technology
2 shared publications
Latest Publications
- Open Access
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- p. 1
Epically Powerful: An open-source software and mechatronics infrastructure for wearable robotic systems
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- p. 1
Open AccessEpically Powerful: An open-source software and mechatronics infrastructure for wearable robotic systems
- 24 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Robotics
- Vol. 41 (15523098) , 4300-4321
Robust-Locomotion-By-Logic: Perturbation-Resilient Bipedal Locomotion via Signal Temporal Logic Guided Model Predictive Control
This study introduces a robust planning framework that utilizes a model predictive control (MPC) approach, enhanced by incorporating signal temporal logic (STL) specifications. This marks the first-ever study to apply STL-guided trajectory optimization for bipedal locomotion, specifically designed to handle both translational and orientational perturbations. Existing recovery strategies often struggle with reasoning complex task logic and evaluating locomotion robustness systematically, making them susceptible to failures caused by inappropriate recovery strategies or lack of robustness. To address these issues, we design an analytical stability metric for bipedal locomotion and quantify this metric using STL specifications, which guide the generation of recovery trajectories to achieve maximum robustness degree. To enable safe and computational-efficient crossed-leg maneuver, we design data-driven self-leg-collision constraints that are 1000 times faster than the traditional inverse-kinematics-based approach. Our framework outperforms a state-of-the-art locomotion controller, a standard MPC without STL, and a linear-temporal-logic-based planner in a high-fidelity dynamic simulation, especially in scenarios involving crossed-leg maneuvers. Additionally, the Cassie bipedal robot achieves robust performance under horizontal and orientational perturbations such as those observed in ship motions. These environments are validated in simulations and deployed on hardware. Furthermore, our proposed method demonstrates versatility on stepping stones and terrain-agnostic features on inclined terrains.
- 24 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Robotics
- Vol. 41 (15523098) , 4300-4321
Robust-Locomotion-By-Logic: Perturbation-Resilient Bipedal Locomotion via Signal Temporal Logic Guided Model Predictive Control
This study introduces a robust planning framework that utilizes a model predictive control (MPC) approach, enhanced by incorporating signal temporal logic (STL) specifications. This marks the first-ever study to apply STL-guided trajectory optimization for bipedal locomotion, specifically designed to handle both translational and orientational perturbations. Existing recovery strategies often struggle with reasoning complex task logic and evaluating locomotion robustness systematically, making them susceptible to failures caused by inappropriate recovery strategies or lack of robustness. To address these issues, we design an analytical stability metric for bipedal locomotion and quantify this metric using STL specifications, which guide the generation of recovery trajectories to achieve maximum robustness degree. To enable safe and computational-efficient crossed-leg maneuver, we design data-driven self-leg-collision constraints that are 1000 times faster than the traditional inverse-kinematics-based approach. Our framework outperforms a state-of-the-art locomotion controller, a standard MPC without STL, and a linear-temporal-logic-based planner in a high-fidelity dynamic simulation, especially in scenarios involving crossed-leg maneuvers. Additionally, the Cassie bipedal robot achieves robust performance under horizontal and orientational perturbations such as those observed in ship motions. These environments are validated in simulations and deployed on hardware. Furthermore, our proposed method demonstrates versatility on stepping stones and terrain-agnostic features on inclined terrains.
- 14 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Medical Robotics and Bionics
- Vol. 7 (2) , 666-677
Machine Learning Enables Rapid Detection of Slips Using a Robotic Hip Exoskeleton
Fall incidents due to slips are some of the most common causes of injuries for industry workers and older adults, motivating research to assist balance recovery following slips. To assist balance recovery during a slip, a detection algorithm that can work with an assistive device, such as an exoskeleton, needs to be able to detect slips rapidly after onset, which remains a critical gap in the field. Here, we compared the ability of linear discriminant analysis (LDA), extreme gradient boosting (XGBoost), and convolutional neural networks (CNN) to detect slip using only native sensors on a hip exoskeleton. We trained and evaluated user-independent models on early-stance (ES) and late-stance (LS) slips of various magnitudes collected through treadmill-based slips. All models, except LDA with LS slips, detected slips with ¿90% accuracy. Overall, the best model was XGBoost, with its fastest results achieving average detection times and median accuracies of 155.06 ms at 96.25% for ES slips and 228.88 ms at 93.75% for LS slips, while also achieving 100% sensitivity at 195.64 ms (ES) and 266.24 ms (LS). Our results indicate a promising direction for further research into designing a generalizable model for balance recovery during slip perturbations using robotic hip exoskeletons.
- 14 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Medical Robotics and Bionics
- Vol. 7 (2) , 666-677
Machine Learning Enables Rapid Detection of Slips Using a Robotic Hip Exoskeleton
Fall incidents due to slips are some of the most common causes of injuries for industry workers and older adults, motivating research to assist balance recovery following slips. To assist balance recovery during a slip, a detection algorithm that can work with an assistive device, such as an exoskeleton, needs to be able to detect slips rapidly after onset, which remains a critical gap in the field. Here, we compared the ability of linear discriminant analysis (LDA), extreme gradient boosting (XGBoost), and convolutional neural networks (CNN) to detect slip using only native sensors on a hip exoskeleton. We trained and evaluated user-independent models on early-stance (ES) and late-stance (LS) slips of various magnitudes collected through treadmill-based slips. All models, except LDA with LS slips, detected slips with ¿90% accuracy. Overall, the best model was XGBoost, with its fastest results achieving average detection times and median accuracies of 155.06 ms at 96.25% for ES slips and 228.88 ms at 93.75% for LS slips, while also achieving 100% sensitivity at 195.64 ms (ES) and 266.24 ms (LS). Our results indicate a promising direction for further research into designing a generalizable model for balance recovery during slip perturbations using robotic hip exoskeletons.
- Open Access
- 1 April 2025
- Published by Human Kinetics in Journal of Applied Biomechanics
- Vol. 41 (2) , 151-160
Electromyography-Informed Estimates of Joint Contact Forces Within the Lower Back and Knee Joints During a Diverse Set of Industry-Relevant Manual Lifting Tasks
Repetitive manual labor tasks involving twisting, bending, and lifting commonly lead to lower back and knee injuries in the workplace. To identify tasks with high injury risk, we recruited N = 9 participants to perform industry-relevant, 2-handed lifts with a 11-kg weight. These included symmetrical/asymmetrical, ascending/descending lifts that varied in start-to-end heights (knee-to-waist and waist-to-shoulder). We used a data-driven musculoskeletal model that combined force and motion data with a muscle activation-informed solver (OpenSim, CEINMS) to estimate 3-dimensional internal joint contact forces (JCFs) in the lower back (L5/S1) and knee. Symmetrical lifting resulted in larger peak JCFs than asymmetrical lifting in both the L5/S1 (+20.2% normal [P < .01], +20.3% shear [P = .001], +20.6% total [P < .01]) and the knee (+39.2% shear [P = .001]), and there were no differences in peak JCFs between ascending versus descending motions. Below-the-waist lifting generated significantly greater JCFs in the L5/S1 and knee than above-the-waist lifts (P < .01). We found a positive correlation between knee and L5/S1 peak total JCFs (R2 = .60, P < .01) across the task space, suggesting motor coordination that favors sharing of load distribution across the trunk and legs during lifting.
- 1 April 2025
- Published by Human Kinetics in Journal of Applied Biomechanics
- Vol. 41 (2) , 151-160
Open AccessElectromyography-Informed Estimates of Joint Contact Forces Within the Lower Back and Knee Joints During a Diverse Set of Industry-Relevant Manual Lifting Tasks
Repetitive manual labor tasks involving twisting, bending, and lifting commonly lead to lower back and knee injuries in the workplace. To identify tasks with high injury risk, we recruited N = 9 participants to perform industry-relevant, 2-handed lifts with a 11-kg weight. These included symmetrical/asymmetrical, ascending/descending lifts that varied in start-to-end heights (knee-to-waist and waist-to-shoulder). We used a data-driven musculoskeletal model that combined force and motion data with a muscle activation-informed solver (OpenSim, CEINMS) to estimate 3-dimensional internal joint contact forces (JCFs) in the lower back (L5/S1) and knee. Symmetrical lifting resulted in larger peak JCFs than asymmetrical lifting in both the L5/S1 (+20.2% normal [P < .01], +20.3% shear [P = .001], +20.6% total [P < .01]) and the knee (+39.2% shear [P = .001]), and there were no differences in peak JCFs between ascending versus descending motions. Below-the-waist lifting generated significantly greater JCFs in the L5/S1 and knee than above-the-waist lifts (P < .01). We found a positive correlation between knee and L5/S1 peak total JCFs (R2 = .60, P < .01) across the task space, suggesting motor coordination that favors sharing of load distribution across the trunk and legs during lifting.
- 1 April 2024
- Published by Springer Nature in Annals of Biomedical Engineering
- Vol. 52 (8) , 2013-2023
A Data-Driven Approach to Estimate Human Center of Mass State During Perturbed Locomotion Using Simulated Wearable Sensors
Center of mass (COM) state, specifically in a local reference frame (i.e., relative to center of pressure), is an important variable for controlling and quantifying bipedal locomotion. However, this metric is not easily attainable in real time during human locomotion experiments. This information could be valuable when controlling wearable robotic exoskeletons, specifically for stability augmentation where knowledge of COM state could enable step placement planners similar to bipedal robots. Here, we explored the ability of simulated wearable sensor-driven models to rapidly estimate COM state during steady state and perturbed walking, spanning delayed estimates (i.e., estimating past state) to anticipated estimates (i.e., estimating future state). We used various simulated inertial measurement unit (IMU) sensor configurations typically found on lower limb exoskeletons and a temporal convolutional network (TCN) model throughout this analysis. We found comparable COM estimation capabilities across hip, knee, and ankle exoskeleton sensor configurations, where device type did not significantly influence error. We also found that anticipating COM state during perturbations induced a significant increase in error proportional to anticipation time. Delaying COM state estimates significantly increased accuracy for velocity estimates but not position estimates. All tested conditions resulted in models with R2 > 0.85, with a majority resulting in R2 > 0.95, emphasizing the viability of this approach. Broadly, this preliminary work using simulated IMUs supports the efficacy of wearable sensor-driven deep learning approaches to provide real-time COM state estimates for lower limb exoskeleton control or other wearable sensor-based applications, such as mobile data collection or use in real-time biofeedback.
- 1 April 2024
- Published by Springer Nature in Annals of Biomedical Engineering
- Vol. 52 (8) , 2013-2023
A Data-Driven Approach to Estimate Human Center of Mass State During Perturbed Locomotion Using Simulated Wearable Sensors
Center of mass (COM) state, specifically in a local reference frame (i.e., relative to center of pressure), is an important variable for controlling and quantifying bipedal locomotion. However, this metric is not easily attainable in real time during human locomotion experiments. This information could be valuable when controlling wearable robotic exoskeletons, specifically for stability augmentation where knowledge of COM state could enable step placement planners similar to bipedal robots. Here, we explored the ability of simulated wearable sensor-driven models to rapidly estimate COM state during steady state and perturbed walking, spanning delayed estimates (i.e., estimating past state) to anticipated estimates (i.e., estimating future state). We used various simulated inertial measurement unit (IMU) sensor configurations typically found on lower limb exoskeletons and a temporal convolutional network (TCN) model throughout this analysis. We found comparable COM estimation capabilities across hip, knee, and ankle exoskeleton sensor configurations, where device type did not significantly influence error. We also found that anticipating COM state during perturbations induced a significant increase in error proportional to anticipation time. Delaying COM state estimates significantly increased accuracy for velocity estimates but not position estimates. All tested conditions resulted in models with R2 > 0.85, with a majority resulting in R2 > 0.95, emphasizing the viability of this approach. Broadly, this preliminary work using simulated IMUs supports the efficacy of wearable sensor-driven deep learning approaches to provide real-time COM state estimates for lower limb exoskeleton control or other wearable sensor-based applications, such as mobile data collection or use in real-time biofeedback.
- 28 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 9 (5) , 3995-4002
Dynamic Duo: Design and Validation of an Autonomous Frontal and Sagittal Actuating Hip Exoskeleton for Balance Modulation During Perturbed Locomotion
Humans are required to maintain balance during locomotion in challenging environments, which present an even bigger challenge for individuals with balance impairments. Exoskeleton-driven balance augmentation is a promising avenue to assist users in these environments, but there has been little work in developing exoskeleton devices for these applications. In this work, we present the design, realization, and validation of an autonomous robotic hip exoskeleton with frontal and sagittal actuation. This device contains four quasi-direct drive-actuated degrees of freedom, enabling both frontal and sagittal assistance at the hip joints. The device is relatively light-weight, weighing 7.76 kg, and low-profile in the frontal plane, enabling unimpeded arm swing of the user. We found that wearing the device did not change walking kinematics, validating that the design does not inhibit the user's natural motion. We validated the exoskeleton using a bilateral bang-bang controller that successfully modulated step width and length in all cardinal and ordinal directions (all p<0.001) during steady state and perturbed walking. We also found that step modulation capability is influenced by swing leg kinematics and perturbation context. Broadly, this work presents a lightweight, autonomous, powered exoskeleton that can be used to study control approaches for balance augmentation
- 28 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 9 (5) , 3995-4002
Dynamic Duo: Design and Validation of an Autonomous Frontal and Sagittal Actuating Hip Exoskeleton for Balance Modulation During Perturbed Locomotion
Humans are required to maintain balance during locomotion in challenging environments, which present an even bigger challenge for individuals with balance impairments. Exoskeleton-driven balance augmentation is a promising avenue to assist users in these environments, but there has been little work in developing exoskeleton devices for these applications. In this work, we present the design, realization, and validation of an autonomous robotic hip exoskeleton with frontal and sagittal actuation. This device contains four quasi-direct drive-actuated degrees of freedom, enabling both frontal and sagittal assistance at the hip joints. The device is relatively light-weight, weighing 7.76 kg, and low-profile in the frontal plane, enabling unimpeded arm swing of the user. We found that wearing the device did not change walking kinematics, validating that the design does not inhibit the user's natural motion. We validated the exoskeleton using a bilateral bang-bang controller that successfully modulated step width and length in all cardinal and ordinal directions (all p<0.001) during steady state and perturbed walking. We also found that step modulation capability is influenced by swing leg kinematics and perturbation context. Broadly, this work presents a lightweight, autonomous, powered exoskeleton that can be used to study control approaches for balance augmentation
- 3 November 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Medical Robotics and Bionics
- Vol. 6 (1) , 224-234
Comparing Metabolic Cost and Muscle Activation for Knee and Back Exoskeletons in Lifting
Performing heavy manual lifting jobs can be extremely strenuous, often resulting in high rates of injury at overloaded joints. Lower-body exoskeletons have potential to mitigate this fatigue and injury through targeted joint assistance. A variety of current wearable devices aim to provide this kind of support, but their effects on both directly and indirectly targeted joints have not been fully investigated. In this study, a powered bilateral knee exoskeleton (research device), and passive back exosuit (HeroWear Apex) were compared across 10 individuals through a battery of manual lifting tasks. Metabolic, electromyographic, and subject-reported outcome measure data were used to determine which exoskeleton provided the best assistance for each task type. It was found that the knee exoskeleton significantly reduced metabolic cost by 9.6% compared to the no-exoskeleton baseline (p<.05). Both the knee and back exoskeletons reduced activation in up to 6 of the muscles measured (out of 16). Importantly, each device primarily reduced muscle activation around the joint it targeted, not around adjacent joints. These findings suggest that exoskeletons that target the knee or lower back joint during lifting can provide significant assistance in reducing exertion.
- 3 November 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Medical Robotics and Bionics
- Vol. 6 (1) , 224-234
Comparing Metabolic Cost and Muscle Activation for Knee and Back Exoskeletons in Lifting
Performing heavy manual lifting jobs can be extremely strenuous, often resulting in high rates of injury at overloaded joints. Lower-body exoskeletons have potential to mitigate this fatigue and injury through targeted joint assistance. A variety of current wearable devices aim to provide this kind of support, but their effects on both directly and indirectly targeted joints have not been fully investigated. In this study, a powered bilateral knee exoskeleton (research device), and passive back exosuit (HeroWear Apex) were compared across 10 individuals through a battery of manual lifting tasks. Metabolic, electromyographic, and subject-reported outcome measure data were used to determine which exoskeleton provided the best assistance for each task type. It was found that the knee exoskeleton significantly reduced metabolic cost by 9.6% compared to the no-exoskeleton baseline (p<.05). Both the knee and back exoskeletons reduced activation in up to 6 of the muscles measured (out of 16). Importantly, each device primarily reduced muscle activation around the joint it targeted, not around adjacent joints. These findings suggest that exoskeletons that target the knee or lower back joint during lifting can provide significant assistance in reducing exertion.
- Open Access
- 15 March 2023
- Published by The Company of Biologists in Journal of Experimental Biology
- Vol. 226 (6)
Linking whole-body angular momentum and step placement during perturbed human walking
Human locomotion is remarkably robust to environmental disturbances. Previous studies have thoroughly investigated how perturbations influence body dynamics and what recovery strategies are used to regain balance. Fewer studies have attempted to establish formal links between balance and the recovery strategies that are executed to regain stability. We hypothesized that there would be a strong relationship between the magnitude of imbalance and recovery strategy during perturbed walking. To test this hypothesis, we applied transient ground surface translations that varied in magnitude, direction and onset time while 11 healthy participants walked on a treadmill. We measured stability using integrated whole-body angular momentum (iWBAM) and recovery strategy using step placement. We found the strongest relationships between iWBAM and step placement in the frontal plane for earlier perturbation onset times in the perturbed step (R2=0.52, 0.50) and later perturbation onset times in the recovery step (R2=0.18, 0.25), while correlations were very weak in the sagittal plane (all R2≤0.13). These findings suggest that iWBAM influences step placement, particularly in the frontal plane, and that this influence is sensitive to perturbation onset time. Lastly, this investigation is accompanied by an open-source dataset to facilitate research on balance and recovery strategies in response to multifactorial ground surface perturbations, including 96 perturbation conditions spanning all combinations of three magnitudes, eight directions and four gait cycle onset times.
- 15 March 2023
- Published by The Company of Biologists in Journal of Experimental Biology
- Vol. 226 (6)
Open AccessLinking whole-body angular momentum and step placement during perturbed human walking
Human locomotion is remarkably robust to environmental disturbances. Previous studies have thoroughly investigated how perturbations influence body dynamics and what recovery strategies are used to regain balance. Fewer studies have attempted to establish formal links between balance and the recovery strategies that are executed to regain stability. We hypothesized that there would be a strong relationship between the magnitude of imbalance and recovery strategy during perturbed walking. To test this hypothesis, we applied transient ground surface translations that varied in magnitude, direction and onset time while 11 healthy participants walked on a treadmill. We measured stability using integrated whole-body angular momentum (iWBAM) and recovery strategy using step placement. We found the strongest relationships between iWBAM and step placement in the frontal plane for earlier perturbation onset times in the perturbed step (R2=0.52, 0.50) and later perturbation onset times in the recovery step (R2=0.18, 0.25), while correlations were very weak in the sagittal plane (all R2≤0.13). These findings suggest that iWBAM influences step placement, particularly in the frontal plane, and that this influence is sensitive to perturbation onset time. Lastly, this investigation is accompanied by an open-source dataset to facilitate research on balance and recovery strategies in response to multifactorial ground surface perturbations, including 96 perturbation conditions spanning all combinations of three magnitudes, eight directions and four gait cycle onset times.
- Open Access
Adapting Semi-Active Prostheses to Real-World Movements: Sensing and Controlling the Dynamic Mean Ankle Moment Arm with a Variable-Stiffness Foot on Ramps and Stairs
(1) Background: Semi-active prosthetic feet can provide adaptation in different circumstances, enabling greater function with less weight and complexity than fully powered prostheses. However, determining how to control semi-active devices is still a challenge. The dynamic mean ankle moment arm (DMAMA) provides a suitable biomechanical metric, as its simplicity matches that of a semi-active device. However, it is unknown how stiffness and locomotion modes affect DMAMA, which is necessary to create closed-loop controllers for semi-active devices. In this work, we develop a method to use only a prosthesis-embedded load sensor to measure DMAMA and classify locomotion modes, with the goal of achieving mode-dependent, closed-loop control of DMAMA using a variable-stiffness prosthesis. We study how stiffness and ground incline affect the DMAMA, and we establish the feasibility of classifying locomotion modes based exclusively on the load sensor. (2) Methods: Human subjects walked on level ground, ramps, and stairs while wearing a variable-stiffness prosthesis in low-, medium-, and high-stiffness settings. We computed DMAMA from sagittal load sensor data and prosthesis geometric measurements. We used linear mixed-effects models to determine subject-independent and subject-dependent sensitivity of DMAMA to incline and stiffness. We also used a machine learning model to classify locomotion modes using only the load sensor. (3) Results: We found a positive linear sensitivity of DMAMA to stiffness on ramps and level ground. Additionally, we found a positive linear sensitivity of DMAMA to ground slope in the low- and medium-stiffness conditions and a negative interaction effect between slope and stiffness. Considerable variability suggests that applications of DMAMA as a control input should look at the running average over several strides. To examine the efficacy of real-time DMAMA-based control systems, we used a machine learning model to classify locomotion modes using only the load sensor. The classifier achieved over 95% accuracy. (4) Conclusions: Based on these findings, DMAMA has potential for use as a closed-loop control input to adapt semi-active prostheses to different locomotion modes.
Open AccessAdapting Semi-Active Prostheses to Real-World Movements: Sensing and Controlling the Dynamic Mean Ankle Moment Arm with a Variable-Stiffness Foot on Ramps and Stairs
(1) Background: Semi-active prosthetic feet can provide adaptation in different circumstances, enabling greater function with less weight and complexity than fully powered prostheses. However, determining how to control semi-active devices is still a challenge. The dynamic mean ankle moment arm (DMAMA) provides a suitable biomechanical metric, as its simplicity matches that of a semi-active device. However, it is unknown how stiffness and locomotion modes affect DMAMA, which is necessary to create closed-loop controllers for semi-active devices. In this work, we develop a method to use only a prosthesis-embedded load sensor to measure DMAMA and classify locomotion modes, with the goal of achieving mode-dependent, closed-loop control of DMAMA using a variable-stiffness prosthesis. We study how stiffness and ground incline affect the DMAMA, and we establish the feasibility of classifying locomotion modes based exclusively on the load sensor. (2) Methods: Human subjects walked on level ground, ramps, and stairs while wearing a variable-stiffness prosthesis in low-, medium-, and high-stiffness settings. We computed DMAMA from sagittal load sensor data and prosthesis geometric measurements. We used linear mixed-effects models to determine subject-independent and subject-dependent sensitivity of DMAMA to incline and stiffness. We also used a machine learning model to classify locomotion modes using only the load sensor. (3) Results: We found a positive linear sensitivity of DMAMA to stiffness on ramps and level ground. Additionally, we found a positive linear sensitivity of DMAMA to ground slope in the low- and medium-stiffness conditions and a negative interaction effect between slope and stiffness. Considerable variability suggests that applications of DMAMA as a control input should look at the running average over several strides. To examine the efficacy of real-time DMAMA-based control systems, we used a machine learning model to classify locomotion modes using only the load sensor. The classifier achieved over 95% accuracy. (4) Conclusions: Based on these findings, DMAMA has potential for use as a closed-loop control input to adapt semi-active prostheses to different locomotion modes.