skip to Main Content

MiRAH Myoelectric Hand

Concept

The idea for the development of MiRAH prosthetic technologies arises from a few simple considerations:

  • several international studies have shown that the use of a prosthesis in pre-school age leads to the internalization of the mechanisms thus allowing a fine and natural control of the prosthetic device, which therefore becomes an integral part of the body and no longer an external element to be controlled “knowingly”
  • an engineering process carried out starting from the small, from the solution for children, then it is easier to replicate on a larger scale, on the contrary the approaches followed starting from an idea born for adults have proved inapplicable even to suitable and suitable solutions for children
  • 3D scanning and printing techniques can help optimize production processes, both in terms of adaptability of the device to the user (having a custom product instead of a standard is a pre-processing and not a production problem) and production costs (new materials and printing techniques have dramatically reduced the costs associated with this part). Last but not least, the possibility of reaching disadvantaged countries through local printing of arts, spare parts or various accessories.
  • A child who has grown up with ‘his’ prosthesis, suitably and constantly adapted to his needs, and supported during development, will become an adult with unprecedented prosthetic maturity.

Starting from these simple considerations, an Innovation driven path is considered practicable which aims not only to innovate in terms of solution but which aims, with a careful eye, on the integration and optimization of the entire process and which allows the creation of modular prosthetic technologies for children.

The key points of the proposal are the following:​

Technology innovation

  • Integration of extremely miniaturized and industrial solutions for the realization of the hand.
  • Microprocessor control system that would potentially allow the development of mapping and control models based on deep-learning and AI.
  • Stall control and grip management.

Process innovation

  • Integration and automation of the 3D scanning and printing process for the construction of the structure.
  • Use of innovative biomedical and biocompatible materials with 3D printing.
  • Modularity and scalability of the product (the modular approach will allow the reuse of the modules with the growth of the patient and the simple replacement of any damaged elements, as well as a high specialization for the different types of use).

Additional features that can be developed

  • Wireless charging and immediate interchangeability of the modules to diversify their use in everyday environments and to have a plug & use approach.
  • Tactile feedback
  • Swarm sEMG sensor array for a more sophisticated and natural movement control.
  • Integration of capacitive type sEMG sensors.
  • Solutions for energy generation (energy harvesting). Omission.

The primary objectives achievable by developing the above issues are:

  • the creation of a modular prosthesis for children;
  • the development of a bionic, modular, sized prosthesis with adequate strength so that it can also be used by adults.

Particular attention and sensitivity is also placed on ethical discourse and on the possibility of making technologies and prosthetic solutions widespread and accessible to the widest audience.

Specifically, it is expected that the development and release will take place in an open source context allowing qualified third parties the production and maintenance of ultra low cost solutions for both adults and children, being able to achieve this type of solutions similar to high end ones currently available.

The Open Source approach would allow transnational governmental bodies, NGOs and foundations to be able to produce or distribute these prostheses at very low cost, making them accessible to an extremely large user base and with a capillary penetration in contexts otherwise inaccessible with other development models and traditional distribution. On the other hand, an open community would be a continuous source of input and indications to rapidly implement improvement and innovations on the developed technologies.

High Level Design​

The high level design of the solution is illustrated below, specifically the overall view and then separately the characteristics of the hardware components, control firmware, EMG sensors, engine, mechanical structure, substructure and enclosure will be presented separately.

Integrated solution

The solution is conceptually made up of some distinct components that integrate properly to achieve the desired functions, a control hardware is provided, this part contains all the elements necessary to power and control the motor and also to process the muscle impulses by transforming them into the appropriate commands. which allow the opening or closing of the hand. The control firmware that runs on the hardware is clearly complementary to the hardware.

TBA: Add scheme

The structure is divided into two macro elements, the components that mechanically implement the “robotic” hand with gripping capacity controlled by the different positions taken by the motor under the direct control of the firmware / software and the ex-structure, which starting from the wrist, houses the hardware components, EMG sensors and the reservoir.

Hardware

As regards the control hardware, it was hypothesized to create it with COTS (Commercial Off-The Shelf) components, limiting the custom construction to only the PCB (Printed Circuit Board) which creates the interconnection between the commercial boards and houses the connectors towards the outside.
This approach allows to minimize production costs (a PCB board with the characteristics described above has costs of a few euros and is easily achievable) and eliminates the need for sophisticated production and assembly processes.


The necessary components are the following:

  1. Battery Charger / Power Supply
  2. 2xLi-Ion Batteries
  3. DC-DC Voltage Regulator
  4. Micro
  5. Motor H-Bridge

The folllowing figure shows the reference architecture.

The blue elements are the COTS cards necessary for the integration of the solution, as can be seen from the simplified diagram, the main interconnections are the power lines, then there are the communication lines between Micro and H-Bridge (the motor driver ) and finally the communication and power lines between the motor and the H-Bridge and between EMG and Micro sensors.

The Custom PCB is shown in green.

As regards the Micro, after various evaluations, the Particle Photon card was chosen, a card based on the STM32 ARM processor with the following characteristics:

  • STM32F205 120Mhz ARM Cortex M3.
  • Broadcom BCM43362 Wi-Fi chip. 802.11b / g / n Wi-Fi.
  • 1MB flash
  • 128KB RAM.
  • On-board RGB status LED (ext. Drive provided)
  • 18 Mixed-signal GPIO and advanced peripherals.
  • Open source design.
  • Real-time operating system (FreeRTOS)
  • Soft AP setup.
  • FCC, CE and IC certified.

The elements that made us opt for this type of system are manifold, specifically the Particle Photon card has a 32Bit processor with high performance (120Mhz), an integrated wifi and a real onboard realtime operating system, this makes it particularly suitable for signal processing and naturally inclined to integrate with external systems (cloud environment), in fact the card is completely manageable from the cloud and has an IDE and an operating system natively integrated for the so-called IOT (Internet Of Things).

For example, it is already supporting the push of information to the cloud, the publication of functions that can always be triggered from the cloud and the OTA update (Over-The-Air, directly from the cloud, without the need for system interventions).

The SoftAP functions make the Wifi connection easy and easy and the use of the wireless network can be fully controlled locally, thus being able to limit its use only to situations where an interconnection with the network is actually required (signal acquisition and device tuning, software update, remote diagnosis), leaving it disabled in regular operation.

Finally, it should be highlighted the contained consumption of the board, which is about 80ma (peak 350ma) with wifi and active cloud connection, stands below 40ma with wifi off (lower than the consumption of an Arduino Uno at 16Mhz, board typically used in hobby projects and which in general does not guarantee stability and performance similar to Particle Photon) and drops to 5 / 7ma with STM in STOP mode and below 1ma with STM in STANDBY mode.

Firmware​

The structure of the control firmware partly reflects the hardware architecture, specifically in the firmware there are drivers for external devices, such as, specifically:

1 – H-Bridge interface and control driver and motor, this software component implements the control interface of the H-Bridge component by means of the I / O signals exposed by it. Export engine control functions:

    • forward motor movement
    • reverse motor movement
    • motor speed control

It then implements the device monitoring functions, recognizing the fault condition (device malfunction), stall (characterized by the exceeding of the current threshold consumption by the motor) and detection and control of the motor position (with consequent blocking of the same if limit positions are reached).

2 – EMG driver

EMG driver, is the software component that analyzes and acquires EMG signals from the sensors via the ADC lines of the micro. These signals are processed, corrected with respect to any offsets, normalized (and integrated, if necessary, by calculating their envelope), so as to numerically represent the impulses that pass through the monitored muscles.

The tuning of the system and the evaluation of the residual signals on the monitored muscle, but also the research and evaluation of the best algorithms and logics of motor and hand control will have to make use of an integrated telemetry system.

This software layer performs data offboarding in a non-invasive way, allows real-time streaming of the information acquired in real time to a target host, which acquires, via the network, the information to historicize it in local archives or process it in a host environment.

The complexity of the planned architecture is amply compensated by the related advantages:

  • it is possible to simply create a rich archive of EMG acquisitions and use them for the evaluation and training of the control algorithms developed
  • the algorithms themselves can be set up in the host environment, in a comfortable and fast way, focusing on the specific problem, without having to manage the complexity of development in an embedded environment
  • the exposure in the cloud environment of the functions of the board also allows you to implement control feedback, thus implementing acquisition and implementation in an embedded environment and control logic in a host environment
  • in a real context of application it is necessary to evaluate the residual muscle functions of the patient’s stump, these software components could, therefore, be the basis of a framework that allows a specialized orthopedic technician to evaluate the correct positioning of the myoelectric sensors

The last element is the control firmware, which represents the glue between all the elements and implements the device management logic, namely analysis and evaluation of the EMG signals, implementation of the control feedback, monitoring and management of the engine by means of the relative drivers and monitored parameters.

The control firmware also controls the device’s online / offline configurability (wifi on / off and Soft AP configuration).

Back To Top