Medical Motion: Actuation System Selection Analysis for Human Assist Applications

Departments - Medical motion


Fluid power actuation systems have significant advantages over comparable electromechanical systems for use in human assistive devices such as prosthetics and orthotics, exhibiting both higher power-to-weight and power-to-volume densities. Furthermore, fluid power systems do not require a geared transmission or consume power to generate holding torque, and can be back-driven.

In this study, a representative duty cycle was chosen to simulate the requirements for normal human knee ambulation. Various hydraulic and pneumatic actuation systems were selected for comparison against comparable electromechanical systems. Each system conformed to the torque requirements while metrics such as size, weight, and cost were minimized. All the major components of each system, from power source to actuator were evaluated, to avoid bias.

The components analyzed were chosen from off-the-shelf selections when possible. In cases when such components were unavailable, specifications were estimated empirically. All the results were based on numerical simulations.

Comparison Metrics
Weight: The weight of the full system is of primary importance in human assist applications. All major components, from power supply to actuator, are included in the weight calculations.

Volume: The total volume (system size) of the entire system must be minimized. Additional size constraints were added such as the cross sectional area of the actuator assembly (less than that of a human knee, 25cm2).

Efficiency: The efficiency was defined as the ratio between output mechanical work and energy consumed from the power supply (battery or CO2 bottle).

Assembly Profile: The form factor of the whole system assembly for prosthetic/orthotic profile design purposes, to increase the usability.

Cost: Off-the-shelf prices were used when possible. In cases when prices were unavailable, best estimates of manufacturing costs were used for a base of comparison.

Power Systems
Six systems were chosen for the comparison, where systems 1 - 4 are battery driven and systems 5, 6 use CO2 bottle as the energy source.

  1. Hydraulic Linear Actuation System
  2. Hydraulic Rotary Actuation System
  3. DC Motor Driven Ball Screw (Linear) System
  4. DC Motor with Gear Box (Rotary) System
  5. Pneumatic Linear Actuation System
  6. Pneumatic Rotary Actuation System (See Image 1 below)

Actuation Kinematics
Rotary Actuation: provides torque that can be directly utilized to actuate the knee joint. Output torque is consistent at all configurations (Systems 2,4,6).

Linear Actuation: provides force that must be converted into torque by acting through a moment arm. Output torque depends on the lever arm configuration (Systems 1,3,5). (See Image 2)

Benchmark Task

The benchmark task was created to represent the normal knee behavior from a typical level ground walking cycle. To simplify the task without losing the fairness for the comparison, two tasks were selected to represent the two extreme conditions of the knee activity: holding a static torque at certain angle (static task) and moving at maximum angular velocity for the full range of motion (dynamic task). In this particular benchmark task, the two tasks were evaluated with the same time duration.

Static Task: Hold at maximum torque at extended end position

Dynamic Task: Actuation with maximum torque and maximum angular velocity moving from one end to the other

Maximum Output Torque: 30Nm

Nominal Angular Velocity: 70 deg/second

Range of Motion: 70°

Duration of Use: 1 hour (including 30min of static task and 30min of dynamic task)


The result contains values for the comparison metrics for each simulated actuation system. The data are shown in bar charts to the right for visual comparison. Summarized results can be found in the following table, where ü represents how the requirements were met. E.g., ü ü ü ü means it is most desirable (e.g. low weight, high efficiency). Desirable characteristics are also highlighted in green while undesirable ones are highlighted in red. (See Images 3 and 4)

Hydraulic systems (Systems 1,2) have advantages of low weight and high efficiency, but have relatively high cost. Leakage and potential injury from high pressure fluid are of also of concern for human assist devices.

Pneumatic systems have the advantages of lower cost and lower working pressure, but have disadvantages in the size and weight of the accompanying power source. However, the energy source (CO2 bottle) has a significant contribution to the weight and the volume of pneumatic systems. If alternative energy source can be used (e.g., H2O2 pressure source or HCCI engine), pneumatic systems can remain competitive.

Electromechanical systems perform well in weight and volume, but are low in efficiency and lack back-drivability, which could be critical in energy harvesting applications.


Yifan Li
PhD Cacndidate, Graduate Assistant
University of Illinois at Urbana Champaign

Elizabeth T. Hsiao-Wecksler phD
Associate Professor
Dept. of Mechanical Science & Engineering
University of Illinois at Urbana Champaign

William K. Durfee

Professor and Director of Design Education
Department of Mechanical Engineering
University of Minnesota

Gino G. Banco, Ph.D.

Technology Development Manager
Human Motion and Control
Parker Hannifin Corporation