slideshow 1 slideshow 2 slideshow 3 slideshow 4 slideshow 5 slideshow 6 slideshow 7 slideshow 8 slideshow 9 slideshow 10 slideshow 11 slideshow 12 slideshow 13 slideshow 14

Se encuentra usted aquí

Inicio » Project

Objetives

Rehabilitation robots design has undergone a huge development in recent years, due to its potential advantages in terms of reduction of repetitive work of physiotherapists, the reduction of costs of treatments, the possibility of monitor physiological variables and the evolution of the patient, or the development of applications for tele-rehabilitation, among others. On the other hand, the ageing of the population may increase interest in the applications of these systems, to increase the demand for treatments in a context of economic constraints.

An important part of the developments in this field has been linked to the exoskeletons as an aid to rehabilitation in neurological injury. Robots, in particular the parallel robots, among which should be noted for arm, foot platforms systems and different systems for rehabilitation of the ankle have deserved less attention. Recently there have proposed other parallel robots for exercises in the knee.

Despite the advantages that a priori offer parallel thefts (lower cost, greater versatility in your application), its use has been limited mainly to the rehabilitation of the ankle, perhaps because rehabilitation loses efficiency as the area of the body of interest away from the body segment on which acts the robotic system. Features that should bring a system based on a parallel robot for diagnosis and rehabilitation of knee, and which therefore constitute the basic problems that must be addressed in its design and development, can be summarized in the following:

  • Cost. Keep it as low as possible both in construction and operation of the system. This suggests the use of mechanical systems the simplest possible (minimum number of degrees of freedom to perform the intended task) and low-cost sensors (for the kinematic data, for example).
  • Usability: Meaning both the ease in operation and adaptation to the clinical staff.
  • Customization. The system capacity to adapt both to the physical characteristics of the patient to treatment and patient. This affects both the mechanical structure of the system, and the control system. It must take into account that current robots have a limited capacity for customization.
  • Safety. So far this problem has been addressed, among other obvious measures (stop mushrooms, etc.) incorporating actuators with extremely limited power, which affects negatively the possible performance of the system as a whole (limited load capacity, movements at very low speeds).
  • Interface with the patient and user. Unlike industrial robots, which are operated by specialists, rehabilitation robots should be handled by clinical staff or even people with functional limitations, in any case without any expertise in this field. Hence have a user-friendly interface system to be a critical but very neglected in the development aspect.

The previous issues/needs are associated with a need for improvements at the scientific level, affecting three different and complementary fields: mechanic, biomechanical and control

MECHANICAL field: Need to modify the mechanical design to improve the safety, efficiency and mechanical reliability, reduce cost, and increase the versatility, which then affects the customization and adaptation to the treatment.

The present proposal is a continuation of the project "methodology of designing systems biomecatronicos. Application to the development of a parallel hybrid robot for diagnosis and rehabilitation. DPI 2013-44227-R ". In the scope of this project has been designed and built a parallel robot with four degrees of freedom with 3UPS-UPR configuration, which is capable of performing the movements established as necessary for diagnosis and rehabilitation of knee tasks: two two rotations and translations. To our knowledge, this robot represents a novelty in the field of parallel robots.

Now, in the course of the project several possibilities of improvement in order have been detected to obtain a really effective rehabilitation system. In the purely mechanical aspects, include, in order of importance, the following:

  • Quality of the work space. A problem that is common in this type of parallel robots is the presence of singularities direct within the workspace of the robot, this is not only a problem for the control, but a problem of unacceptable security. Taking into account that in this case the trajectories are given by the rehabilitation needs, a possible solution would be to evolve from the current design to one reconfigurable. The topology of the same, which ensures the required movements, would remain and it would modify certain parameters of the same, for example, the location of the junction points of fixed and mobile platforms legs of the robot. All this leads to a problem of optimization that would not only eliminate the singularities of the robot, but also improve dexterity on the type of paths that you must be.
  • Clearances. At this point you must combine that it is looking for a low cost robot with the necessary precision in positioning, given the implications involved in the safety of the user. In the East robot parallel, is the central leg RPU responsible for imposing restrictions on movement, which the study of the influence of clearances in the pairs kinematic positioning of the mobile platform, will mainly focus on the same.
  • Parasitic movements. Always thinking of the safety of the user, it is necessary to detect the possible presence of unanticipated movements, the implications in terms of safety may have and if the way of avoiding them.

 

BIOMECHANICAL field: Need to have models customizable biomechanical allowing to monitor physiological variables needed for control, security, adapting to the treatment:

The development of new devices applied to rehabilitation has gone in parallel with improvements in hardware, sensors and also in control schemes, which have evolved from position control systems, own passive exercises, up to [16] Adaptive control systems. However, the improvement in the efficiency of rehabilitation robotic systems does not depend on only the Mechatronics system design, but that it is necessary to deepen the interaction between the human body and the rehabilitation device, also starting from models that allow you to analyze efforts transmitted by the robot to different body structures. There are several reasons for this:

  • Safety reasons, and that knowledge of efforts is essential to avoid uncomfortable or harmful for patients.
  • The monitoring of efforts is required to develop certain types of exercises, such as the isotonic [16.20], in the development of adaptive control systems Assist-as-Needed, and hybrid systems with electro-stimulation control.
  • The joint analysis of the human robot-body system, can make important improvements in the design of the robot, to consider the response body as a design goal.

 

The estimation of muscle or joint efforts can be made with different levels of complexity and using information of different nature. With respect to the level of complexity is they can use simple models that only analyse the global actions (forces and moments) either down to the detail of the actions in muscles, tendons and ligaments. There are arguments for and against the two approaches. On the one hand, simple models are easier to compute in real time, they can be more easily adapted to the characteristics of the patient and does not present the problem of the indeterminacy of muscular efforts that appears in more detailed models. Against, they do not provide detail information about actions in tendons and ligaments, which may be necessary to control the robot assistance or determine conditions of security. More elaborate models do offer this information, but at the expense of numerous hard-to-customize parameter. Given the sensitivity of the actions joint to small errors, it is possible that the estimates of the forces are of doubtful validity. On the other hand, the indeterminacy of muscle actions forces to define optimal criteria that also affect the final result.

As regards the type of input information, can handle only a dynamic reverse model powered by the movement and the information from sensors on the robot-patient interface or monitor EMG activity and estimate the efforts by a suitable model. Although there is a wide research activity aimed at the development of systems with EMG, the greater level of detail of these models involves much more experimental complex and also a large number of assumptions that may be subtracted validity to this type of estimates. Inverse Dynamics models can offer global information sufficient in some applications, for example to analyze moments joint, either much more elaborate models to define muscle activity.

Despite all this research activity, remain many problems to be resolved in the parallel development of models applicable to robots for lower limb. Almost all activity related to the integration of models focuses on exoskeletons, where there is a direct relationship between the robotic device and the body segment. The application to parallel robots for lower limb is much smaller, especially in the case of parallel robots for exercises in the knee. In these cases, the models are much more complex and with more indeterminations, a strategy based on simple inverse Dynamics models being more realistic in that estimate is only stocks joints. Finally, how to solve the customization of the model in the case of more complex models in which the number of parameters can be large and significantly affects the end result seems not clear. An alternative to this contradictory situation, detail vs. validity, can be to find a compromise solution in which intermediate models which presents a simplified model for reactions in the joints and in the major muscles are used using a static optimization. The model is easily customizable to adapt it to the personal characteristics of each subject, critical aspect in clinical applications.

 
 
CONTROL ENGINEERING field: Need to develop advanced control strategies of parallel robot to perform different exercises and rehabilitation strategies.

Conventional rehabilitation strategies are divided into three groups: compensatory approaches (training patients to use the undamaged body segments to perform tasks before Members made if affected), approaches neurofacilitadores (focusing on rejuvenation of lost motor skills) and repetitive approaches the task (to try to drive to the reorganization of the primary motor cortex of adult through the daily practice of motor activities specific). For this reason the scope of control is determined by the therapeutic approach that will be used. To this day, there is also evidence of the influence of a) therapeutic intensity, b) the active participation of the patient, and c) the error in the position (if there is no error, no motor learning).

It is necessary to note that, while the number of therapeutic approaches is relatively high, the application to parallel robots for rehabilitation of lower limb is much smaller. Tackling techniques of rehabilitation by means of parallel robots, the strategies that most have been developed and applied are:

  • Position control.
  • Control of force and impedance control.
  • Electromyography sensors-based control.
  • Adaptive control.
In the previous draft (DPI 2013-44227-R (see https://mebiomec.ai2.upv.es/ ), held different position for a parallel robot controls based on the dynamic model so that the response of the robot had good performance, with movements accurate as with demanding behaviors. Also developed an Adaptive controller which was carried out in real time an estimate of parameters that are unknown or that can change in a meaningful way.

In addition to position controls were drivers of force through the use of an industrial sensor of 6 degrees of freedom. Thanks to these you could control the interaction between the mobile robot platform and environment (the leg of the patients). An extension of this work resulted in the development of hybrid drivers position/strength that allowed you to specify the robot motion and force references.

                

 

                                       

Theme by Danetsoft and Danang Probo Sayekti inspired by Maksimer