Every 20 seconds, a person with diabetes has to face an amputation somewhere in the world. The main risk factor is high plantar pressure combined with foot insensitivity. In this project, the main goal is to develop a intelligent device in the shape of a shoe that can autonomously redistribute and offload peak pressure points in the foot.

Shape Memory Alloy (SMA) actuators, when using temperature as input, are challenging to control due to SMA's no-linear relationship with displacement and force output, often requiring compensation for effective control. In the literature, SMA models have been proposed a way to provide this compensation, but this method has its drawbacks. The non-linear output behavior of SMA changes due to various initial and state-dependent variables such as temperature, stress or strain, pre-stress, martensite fraction, alloy composition, etc., so the compensation may become inaccurate if any of those variables are not taken into account by the model and measured. An approach that could compensate for the non-linearity of SMA without relying on its initial and state-dependent variables would be more desirable.

Robotic applications in harsh environments face challenges related to the incompatibility of their actuator with the operating conditions. Shielding can reduce the risk of failure, but a more effective way could be the usage of electronic-less actuators.

Wet SMA (Shape Memory Alloy) actuators can be utilized in electronic-less robots to enable motion in harsh environments. These actuators can be designed to work in challenging conditions where traditional electronic-based actuators may fail, such as in extremely high magnetic fields, high humidity, corrosive or explosive environments, or underwater environments. Wet SMA actuators are made from special alloys that exhibit shape memory behavior, meaning they can change shape in response to external stimuli, such as temperature changes or electrical currents, and then return to their original shape when the stimulus is removed. 

Multi-element SMA actuators produce a nonuniform stress distribution, related to the length difference between elements. For the purpose of mitigating the effects of this length difference in the actuator's output, we proposed the use of a differential wire mechanism that joins the wire bundles in antagonistic pairs that auto-compensates for this length difference. 

The proposed mechanism was tested for different prestresses and combinations of the number of wire bundles in parallel to determine if its scaling factor or maximum output force suffers any variation. 

We observed that the scaling factor and the maximum output force increase when using the differential mechanism, as compared to simply assembling the wire bundles directly, an effect more prominent with lower prestress. We also observed that if the differential mechanism is used, the scaling factor does not significantly vary between the prestresses that we tested.

Here, the concept of scalability for actuators was introduced and explored, which is the capability to freely change the output characteristics on demand: displacement and force for a linear actuator, angular position and torque for a rotational actuator. This change can either be used to obtain power improvement (with a constant scale factor) or to improve the usability of a robotic system according to variable conditions (with a variable scale factor). 

Some advantages of a scalable design include the ability to adapt to changing environments, variable resolution of step size, ability to produce designs that are adequate for restricted spaces or that require strict energy efficiency, and intrinsically safe systems. Current approaches for scalability in actuators have shortcomings: the method to achieve scalability is complex, so obtaining a variable scaling factor is challenging, or they cannot scale both output characteristics simultaneously.

Shape Memory Alloy (SMA) wire-based actuators can overcome these limitations, because its two output characteristics, displacement and force, are physically independent of each other. We presented a novel design concept for linear scalable actuators that overcome SMA design and scalability limitations by using a variable number of SMA wires mechanically in parallel, immersed in a liquid that transmits heat from a separate heat source (wet activation). In this concept, more wires increase the maximum attainable force, and longer wires increase the maximum displacement. 

We constructed and tested prototypes with a different number of SMA wires to determine force vs. temperature behavior and time response. The heat-transmitting liquid was either static or flowing using pumps. Scalability was achieved with a simple method in all tested prototypes with a linear correlation of maximum force to the number of SMA wires. With our experiment conditions, flowing heat transmission achieved higher actuation bandwidth than static.

Initially, we presented the basic performance evaluation of the design concept for linear scalable Shape Memory Alloy (SMA) wire actuators controlled by the temperature of a liquid. Fluid heat transmission with SMA wires, contrasting with Joule heating, allows the design to provide scalability of its maximum output force and displacement by changing the arrangement of the wires. The purpose of this design concept is to have an actuator that can scale output energy with minimum changes in footprint or power source.

We did isometric tests on prototypes using this design concept to contrast their force vs. temperature response, as well as their thermal response time. All prototypes successfully achieved force output scalability by changing the number of the parallel actuating SMA wires in a range from ten to forty. In addition, with the addition of flow heat transmission, the force rise time was reduced by 92%, and the force fall time was reduced by 95%, as compared with a design with only nonflowing heat transmission. 

This design concept allows an actuator to be adapted to changing specifications by customizing its output characteristics and could potentially provide further flexibility to robotic systems.

Gloves and similar devices currently used for human-machine force feedback in virtual reality environments have gained increased popularity and multiple applications, from videogames to real estate and medicine. It was the purpose of this project to analyze the feasibility of a new design approach for force feedback using a Shape Memory Alloy (SMA) wire (nitinol) as the main actuator. The actuator was own-designed and integrated with a flexible glove first, and then with a rigid glove, similar to an exoskeleton, which in turn were complemented with the design of a power stage, a control stage to allow multiple levels of response and finally, a computer interface programmed using NI Labview. Also, the contrast between using SMA vs. traditional (e.g electromagnetic) actuators for human virtual reality force feedback were discussed.