Design Innovation Paper

A Robotic Ankle–Foot Prosthesis With Active Alignment

[+] Author and Article Information
Andrew Kennedy LaPrè

Department of Mechanical and Industrial Engineering,
University of Massachusetts Amherst,
160 Governors Drive,
Amherst, MA 01003
e-mail: alapree@umass.edu

Brian R. Umberger

Department of Kinesiology,
University of Massachusetts Amherst,
30 Eastman Lane,
Amherst, MA 01003
e-mail: umberger@umass.edu

Frank C. Sup, IV

Department of Mechanical and Industrial Engineering,
University of Massachusetts Amherst,
160 Governors Drive,
Amherst, MA 01003
e-mail: sup@umass.edu

1Corresponding author.

Manuscript received November 18, 2014; final manuscript received February 10, 2016; published online May 12, 2016. Assoc. Editor: Venketesh Dubey.

J. Med. Devices 10(2), 025001 (May 12, 2016) (9 pages) Paper No: MED-14-1268; doi: 10.1115/1.4032866 History: Received November 18, 2014; Revised February 10, 2016

An ankle–foot prosthesis designed to mimic the missing physiological limb generates a large sagittal moment during push off which must be transferred to the residual limb through the socket connection. The large moment is correlated with high internal socket pressures that are often a source of discomfort for the person with amputation, limiting prosthesis use. In this paper, the concept of active alignment is developed. Active alignment realigns the affected residual limb toward the center of pressure (CoP) during stance. During gait, the prosthesis configuration changes to shorten the moment arm between the ground reaction force (GRF) and the residual limb. This reduces the peak moment transferred through the socket interface during late stance. A tethered robotic ankle prosthesis has been developed, and evaluation results are presented for active alignment during normal walking in a laboratory setting. Preliminary testing was performed with a subject without amputation walking with able-bodied adapters at a constant speed. The results show a 33% reduction in the peak resultant moment transferred at the socket limb interface.

Copyright © 2016 by ASME
Your Session has timed out. Please sign back in to continue.



Grahic Jump Location
Fig. 1

The gastrocnemius and soleus muscles in a person with amputation generate a majority of the sagittal moment about the ankle, but also counter the moment transfer through bone tissue, keeping the tibia mostly in compression (a). High peak pressures are observed on the patellar tendon and distal posterior regions of the residual limb when using a conventional prosthesis as a result of high moment transfer through the socket interface (b). The active alignment prosthesis realigns the residual limb toward the CoP during midstance to reduce moment transfer while producing net-positive work (c).

Grahic Jump Location
Fig. 2

Active alignment prosthesis tethered prototype shown with protective covers and foot shell. The prototype was developed to evaluate modified gait mechanics for reduced moment loading of the residual limb and restoration of rest of body biomechanics.

Grahic Jump Location
Fig. 3

Solid model of the active alignment prosthesis. The design is shown without covers, featuring major components (a). The prosthesis model is displayed in a neutral position (b) and fully extended (c) showing the modified kinematics.

Grahic Jump Location
Fig. 4

Four-bar linkage design parameters with optimal values. The linkage motion directions xfoot, yfoot, and θfoot are shown in the shank reference frame as commonly modeled in practice (proximal to distal). The foot motion can be calculated as a function of a variable linear actuator length and the given design parameters. The optimal link lengths given are the distances between joint centers on the links indicated in subscript. The optimal rigid attachment angles, of links 1 and 3 to the shank and foot, respectively, are given.

Grahic Jump Location
Fig. 5

The single axis moment sensor is shown with strain gages on the underside (a) and the custom sensor PCB (b), which amplifies the moment sensor signal and measures inertial dynamics

Grahic Jump Location
Fig. 6

Experimental setups used for characterization and initial testing. To characterize the moment sensor, the bottom was bolted to ground and a 0.25 m moment arm was attached to the top to apply known moments (a). For tuning and step/frequency responses, the prosthesis was mounted rigidly to ground by the moment sensor and pinned to ground at the foot connection for zero moment loading in four increments of extension (b). For power output testing, the prosthesis was mounted in a custom dynamometer that bolts the moment sensor to ground and pins the foot connection to a spring in series with a load cell and to a linear potentiometer in parallel to the spring and load cell (c).

Grahic Jump Location
Fig. 7

Experimental prosthesis able-bodied adapters allow a person without amputation to walk on the prosthesis for preliminary testing. The contralateral limb is attached to a pylon and matching passive ESAR foot prosthesis aligned to match the neutral (retracted) position of the experimental prosthesis. The use of adapters accelerates the controller development and tuning without the need for test subjects with amputation.

Grahic Jump Location
Fig. 8

Flow diagram of control methodology. A finite state controller is used to schedule gains for a PID position controller for each state. A reference target is calculated for each state based on cadence, state, and the stance time of the previous step.

Grahic Jump Location
Fig. 9

Moment sensor accuracy showing the known applied moment in comparison to the moment recorded by the moment sensor and the RMS deviation as a shaded region

Grahic Jump Location
Fig. 10

Moment generated at the moment sensor of the prosthesis when fixed extended 66% in a bench top jig and commanded a maximum current to the motor. Average rise times at all the positions were less than 60 ms for 90% final value, having standard deviations that overlapped.

Grahic Jump Location
Fig. 11

Closed-loop force control frequency response of the prosthesis extended 66%. Fixed in a bench top characterization apparatus, the system was very stiff and did not display magnitude cutoff. The prosthesis was observed to be controllable up to about 20 Hz. Shortly after resonating at about 20 Hz, the response exhibited higher modalities and became unstable. Responses in all the amounts of extension were similar, having standard deviations that overlapped with the response shown.

Grahic Jump Location
Fig. 12

Comparison of moments recorded by the prosthesis moment sensor while walking on a treadmill at a constant speed of 1.0 m/s while the prosthesis is regulating at a neutral alignment and actively aligning. The lines are averaged values for ten consecutive steps, and the shaded areas are standard deviations. The recorded data show a peak moment reduction of 33% during active alignment.




Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In