Secondary Logo

Journal Logo


The FLUIDHAND III: A Multifunctional Prosthetic Hand

Gaiser, Immanuel Nicolas Dipl.-Ing; Pylatiuk, Christian Dr. med; Schulz, Stefan Dr. Ing.; Kargov, Artem Dr. Ing.; Oberle, Reinhold Dipl.-Math; Werner, Tino Dipl.-Ing

Author Information
JPO Journal of Prosthetics and Orthotics: April 2009 - Volume 21 - Issue 2 - p 91-96
doi: 10.1097/JPO.0b013e3181a1ca54
  • Free

“The soul of men can be described best, through the movements and actions of the body.” This statement by Leonardo da Vinci describes the value of the human hand from a semiotic point of view which is ancillary to the functional properties of a hand prosthesis.1 In light of this, complex hand prostheses have practical and emotional impact on the patient.

Vocational retraining and reentry is the most important thing for most patients after an upper-limb amputation. The goal is to recover as many skills as possible.

The desire of prosthetic hand users for improved functionality was expressed in several surveys.2–4

Patient surveys help to clarify the requirements for upper-limb prostheses. Thus, the development process at the Karlsruhe Institute of Technology KIT (formerly “Forschungszentrum Karlsruhe”) started with a detailed patient survey. The intention was to define user requirements that express the needs of upper-limb prosthetic users.5 The results show clearly three main requests:

  1. Weight reduction of the prosthesis.
  2. Enhancement of the anthropomorphic appearance and the cosmetics.
  3. Extension of the functionality to provide different grasping patterns.

Commercially available hand prostheses mostly provide only one active grasping pattern. An exception is the i-Limb hand produced by Touch Bionics, which is the only prosthetic hand on the market that can show various finger positions.6,7 The various grasping patterns can be achieved by blocking or releasing the different fingers with the other hand or the use of the patient's environment. When comparing the results of the patient survey with the market of upper-limb prostheses, there is still room for improvement.

Contrary to the current research in lower-limb prostheses, there is little current emphasis in developing hydraulics in hand prosthetics.8 The development of pneumatically driven prosthetic hands and arms started in 1948 at the Orthopaedic Hospital in Heidelberg.9 Providing a unique mechanism, the “Heidelberg Hand,” was state-of-the-art in the 60s, but lost importance because of the enhancements in DC motor technology. Also its bulky appearance, complicated recharging, and handling forced patients to use their residual limb or their feet rather than the prosthetic hand. The reemergence of fluidic devices requires the availability of new materials, production technologies, and miniaturized pumps, valves, and actuators.


One of the biggest challenges when designing hand prostheses is to fit everything that is necessary to operate the hand in the metacarpus of the hand. Thus, it can be guaranteed that patients with long residual forelimbs also can be treated. The only components that are situated in the socket are the battery and myoelectrodes. Fluidic actuation systems are able to fulfill the required power density.10 They can be characterized through three main units: pressure generation, pressure distribution, and pressure transformation. The three extraordinary units that activate the FLUIDHAND III are described in the following three sections. An overview of the hand as well as the integration of the operation units are shown in Figures 1A, B, respectively.

Figure 1.:
Overview of the FLUIDHAND III (A), operation units of the hydraulic system (B).


When looking at the characteristics of different pump designs, the best relationship between flow rate, maximum pressure, weight, and size has to be found. After evaluating the different pump designs, external gear pumps (Figure 2) were chosen as meeting the requirements best. This gear pump is designed to provide the best performance possible to operate a prosthetic hand and was described by Kargov et al.11 in 2007. The relevant issues are:

Figure 2.:
Exploded view of the external gear pump.
  • High-flow rate (625 mL/min) → high grasping speed (closing time of hand 1s).
  • High maximum pressure (9.8 bar) → high holding force (holding force of hand 65N).
  • High efficiency (28%) → long operation time.
  • Small dimensions (fits in the metacarpus of a hand) → large variety of achievable hand sizes.
  • Low-noise emissions (55 dB at 1-m distance).
  • Actuation fluid is water → environmentally and maintenance friendly fluid.


The specifications for a valve used in a prosthetic hand cannot be achieved with standard valves. The capability of handling high pressure, allowing a high-flow rate, and being compact and lightweight at the same time are the challenges in this case. The development of tailored valve components is necessary. After an extensive evaluation process, cartridge valves were found to have the best potential.11 Based on the cartridge principle, new valves were developed. The valves are optimized concerning size, flow rate, efficiency, liquid resistance, and pressure stability. The valve demands a high precision manufacturing process. Little irregularities have a tremendous impact on the operability of the valve. Figure 3 displays the standard valve at the left and the custom valve at the right. With the developed valves a well-balanced ratio between dynamics, precise and effective controllability, weight, and size can be achieved.

Figure 3.:
Commercial valve (left) versus custom valve (right).


After creating and controlling the pressurized fluid, the stored energy needs to be transformed into the rotary movement of the single finger joints. Flexible fluidic actuators have the capability to transform pressure directly into rotation.12 Hence, these types of actuators are a precise match to move the single joints of an artificial hand. After an intensive development process, these actuators feature outstanding properties. In addition to the very good power to weight ratio and high-pressure stability, they provide inherent compliant actuation of the finger joints at a maximum torque of 3.7 Nm. When varying the diameter of the actuators they can be adjusted to the joint they are driving or the patient's hand size. Two different sizes of actuators are shown in Figure 4.

Figure 4.:
A 20 mm actuator (top), 12 mm actuator (bottom).

A flexible fluidic actuator principally consists of flexible bellows. These bellows, once they are inflated, apply a torque to the joint and the finger moves. The innovation of these bellows lies in the configuration of different materials. The configuration combines natural rubber, fiber preforms, and high-strength aluminum. This setup guarantees a high number of life cycles and high-pressure stability.

The torque decreases depending on the angle of finger flexion. This effect can be compensated by higher pressure in the fluidic system.


The previous section described the main technical components of the FLUIDHAND III. Although essential for the reliable use of the FLUIDHAND, the technical parts are only of interest to the patient in terms of controllability. When putting on the FLUIDHAND, the first thing to know is how to control it. The practical approach of how to control five grasping patterns with two myoelectrodes is the topic of this section and has been described by Reischl et al.13–15 In addition to standard components such as myoelectrodes and a battery, a controller board is needed to operate the pump and valves in the desired way.

The developed controller board combines a PIC24H256 microcontroller, valve and pump drivers as well as a Bluetooth adapter.

Generally the controller board allows vector (multifunctional) control and signal (Morse) control. Both ways are practical approaches; for one patient vector control might be suitable and for another one, signal control might be the right choice.

The powerful microcontroller board provides the functionality to adjust the signal codes and vector codes to suit the patient's possibilities.

The signal code can be understood as a Morse code in which a grasping pattern is represented by a characteristic signal combination of two myoelectrodes on antagonistic muscles. However, the controller board offers the possibility of connecting different signal codes with different grasping patterns according to frequent tasks of the patient or his or her capabilities to provide myosignals.

Although it is the less of a physiological controlling approach, clinical trials have shown that patients can implement claimed grasping patterns quickly and reproducibly.12

Vector control allows for a more intuitive use of the prosthetic hand. However, this approach requires a good repeatability of the myosignals. The patient has to complete calibration period of about 15–30 minutes. During that time the myoelectrodes are scanned with a frequency of 100 Hz while the patient is performing repetitions of five natural movements (virtual movements). Each natural movement represents the switch signal for one of the five grasping patterns. All test patients were able to operate the prostheses in the desired way with a signal reproduction reliability of 90.4%–99.5%.13–15


Another outstanding specification of the FLUIDHAND III is the possibility to process force feedback signals. To obtain force signals, a force resistive sensor is integrated in the index finger and the thumb, respectively. These signals are then fed in the microcontroller. A vibration motor is connected to the system and, depending on the signal strength, the degree of vibration changes. Patient tests have shown that a force-feedback system has a wide variety of effects on the user.16,17 First of all, it allows the user to regulate the grasping force in relation to the force required. This is especially of interest when grasping sensitive things, persons, or objects without having eye contact. This saves energy and thus increases the time before recharging the battery.


The integration of a Bluetooth Adapter in the controller unit widens the range of possibilities in several ways:

  • Wireless operation of the prosthesis without using myosignals. Easy for testing demonstration and maintenance.
  • Wireless programming of the prosthesis to customize the hand to the patient's needs.
  • In connection with a PDA service, tasks can be accomplished without the need to travel or to leave the personal environment.
  • Training and entertainment can be connected because the prosthesis can serve as a “Bluetooth-Gamepad” to control simple computer games and train prosthesis skills at the same time.


The FLUIDHAND III is a further development of FLUIDHANDs presented by Schulz et al.18 The FLUIDHAND III combines enhancements in functionality, materials, design, controlling, and patient adaptation. Some pictures taken during a clinical trial are shown in Figure 5.

Figure 5.:
Six different hand postures and grasping patterns achievable (A). Natural appearance of the FLUIDHAND III (B and C).

Comparing the grasping forces of different prostheses depends to a large extend on the adaptiveness of grasping. The contact area between hand and object is the determining factor. The required grasping forces for a secure grasp decrease with an increase of the contact area. Grasping studies showed that the human hand needs comparatively low grasping forces for most daily life tasks because of its ability to adapt.19

To determine the pure grasping force, several tests were conducted.

In a first test setup, a PMMA cylinder with a diameter of 60 mm is attached coaxially to a force gauge. The cylinder is grasped with the prosthesis and then a force is applied as shown in Figure 6 (left). The maximum grasping force before the cylinder starts to slip was determined with 65 N. This value can be defined as holding force.

Figure 6.:
Holding force test (left), fingertip grasping force test (right).

The second test setup depicts the grasping force in a single fingertip. The fingertip is attached to the force gauge as shown in Figure 6 (right). The maximum measured force was 45 N. The force characteristics depend on the angle of the proximal finger joint. This value can be determined as the fingertip grasping force. The fingertip grasping force lies in between 45 N at an angel of 0° and 5 N at an angle of 90°.

In addition to the tests mentioned above a testing environment for artificial hands the “Artificial Patient” was developed. This test rig permits the appliance of defined forces in direction and value to the artificial hand. Simultaneously, the compliance of the hand can be measured in different directions by measuring the deflection of each actuator of the test rig. The “Artificial Patient” provides the opportunity to apply lifecycle tests, maximum stress tests, and daily life tests. The testing environment completes the preclinical testing and evaluation. An overview of the test rig is shown in Figure 7.

Figure 7.:
Artificial patient.

The test setups described try to characterize the grasping and holding force from a practical point of view, always focusing on the benefit for the patient. The complete specifications of the FLUIDHAND III are summarized in Table 1.

Table 1:
FLUIDHAND III specifications


As described in the previous sections, fluidic actuators exhibit a very good power to weight ratio. That means the power density increases with every additional actuator because only one pressure supply is needed. In the prosthetic and rehabilitation context, this fact expands the field of application toward hand-arm solutions. Bigger and stronger fluidic drive elements have been developed and evaluated in earlier projects.20Figure 8 shows a conceptual study of a fluidic hand-elbow solution.

Figure 8.:
Fluidic Hand—elbow combination.


Flexible fluidic actuators have been shown to be a very suitable actuation system for artificial hands and hand prostheses. The outstanding power to weight ratio in combination with the inherent compliance of the actuators meet the requirements of a hand prosthesis very well. Further research will include the enhancement of each component of the actuation system as well as the whole hand performance. It is desired to establish serial production. Another focus lays on developing a hand that provides reliable use even in extreme environmental conditions.


The authors thank the medical staff and technicians from the Orthopaedic University Hospital, Heidelberg, Germany; Pohlig Orthopädietechnik, Traunstein, Germany; and Brillinger Orthopädietechnik, Tübingen, Germany, for their support and help.


1.MacCurdy E. The Notebooks of Leonardo da Vinci. London, UK: Jonathan Cape; 1977.
2.Silcox DH, Rooks MD, Vogel RR, et al. A long-term follow-up and a study of the use of alternate prostheses. J Bone Joint Surg 1993;75:1781–1789.
3.Pylatiuk C, Kargov A, Werner T, et al. Entwicklung einer multifunktionalen Handprothese für Kinder. BMT 2007-41. Jahrestagung der Deutschen Gesellschaft fürBiomedizinische Technik im VDE. Aachen, Germany, September 26–29, 2007.
4.Kyberd PJ, Wartenberg C, Sandsjö L, et al. Survey of upper extremity prosthesis users in Sweden and the United Kingdom. J Prosth Ortho 2007;19:55–62.
5.Pylatiuk C, Schulz S, Döderlein L. Results of an internet survey of myoelectric prosthetic hand users. J Prosth Ortho 2007;31:362–370.
6.Newman PJ. Next generation bionic hand technology—an independently controlled test of the i-limb hand. Proceedings of the 12th World Congress of theInternational Society for Prosthetics and Orthotics. Vancouver, Canada, July 29–August 3; 2007.
7.Newman PJ. Multi-articulating hands & fingers—technical strategies for improved function and myoelectric uptake. Proceedings of MEC '02, MyoElectric Controls/Powered Prosthetics Symposium. Fredericton, Canada, August 13–15, 2008.
8.Puchhammer GE. Future actuating technologies for upper limb extremity prosthetic devices. In: 10th International Conference on New Actuators ACTUATOR 2006. Bremen, Germany, June 14–16, 2006.
9.Marquart E. The Heidelberg pneumatic arm prosthesis. J Bone Joint Surg 1965;47:425–434.
10.Hollerbach JM, Hunter IW, Ballantyne J. A comparative analysis of actuator technologies for robotics. In: Khatib O, Craig J, Lozano-Perez T, eds. The RoboticsReview 2. Cambridge, MA: MIT Press; 1992:299–342.
11.Kargov A, Pylatiuk C, Schulz S. Development of a miniaturised hydraulic actuation system for artificial hands. Sensors Actuators A: Phys 2007;141:548–557.
12.Schulz S. A New Adaptive Hand Prosthesis Based on Flexible Fluidic Actuators [Dissertation]. Karlsruhe: Department of Mechanical Engineering, University of Karlsruhe (KIT); 2004.
13.Reischl M, Mikut R, Pylatiuk C, et al. Control and signal processing concepts for a multifunctional hand prosthesis. In: Proceedings of MEC ‘02, MyoElectric Controls/Powered Prosthetics Symposium. Fredericton, Canada, August 21–23, 2002.
14.Reischl M, Pylatiuk C, Mikut R. Individual control concepts for multifunctional hand prostheses. Orthopädie Technik (English Edition) 2005;I:12–14.
15.Reischl M, Mikut R, Bretthauer G. Robust training and control strategies fort he grasp type selection of hand prostheses. In: Proceedings of the IFAC Symposium onMechatronic Systems. Heidelberg, Germany, September 12–14, 2006.
16.Pylatiuk C, Kargov A, Schulz S, et al. A sensory feedback system for prosthetic hands. In: 14th International Conference of Medical Physics (ICMP 2005). Nürnberg, Germany, September 14–17, 2005.
17.Pylatiuk C, Kargov A, Schulz S. Design and evaluation of a low-cost force feedback system for myoelectric prosthetic hands. JPO J Prosth Ortho 2006;18:362–370.
18.Schulz S, Pylatiuk C, Reischl M, et al. A hydraulically driven multifunctional prosthetic hand. Robotica 2005;23:293–299.
19.Kargov A, Pylatiuk C, Martin J, et al. A comparison of the grip force distribution in natural hands and in prosthetic hands. Disability Rehabil 2008;26:705–711.
20.Kargov A, Breitwieser H, Klosek H, et al. Design of a modular arm robot system based on flexible fluidic drive elements. In: Proceedings of the 2007 IEEE 10thInternational Conference on Rehabilitation Robotics. Noordwijk, The Netherlands, June 12–15, 2007.

prosthetic hand; adaptive grasping; force feedback; grasp stability; patient-specific control parameters; multifunctional prosthesis

© 2009 American Academy of Orthotists & Prosthetists