Summary Analysis Draft 4

Soft Robotic Arm uses flexible sensors to understand its position 

MEC 1281

Summary Analysis

Draft #4

By Sim Wei Yan

4th Feb 2021

In the article, “Soft robotic arm...” (2020), Matheson described Ryan Truby, a postdoc in the MIT Computer Science and Artificial Laboratory, empowering a malleable robotic arm to operate by utilizing motion and data in 3D environment with its own sensorized skin. Matheson mentioned Truby's opinion that soft robot are more advantageous compared to traditional rigid design. This is due to their infinite number of movements at any time and how the soft robot uses their own flexible sensors for control instead of vision system to provide feedback. However, this creates limitation in the control application and planning process due to the infinite numbers of movement. Using of the soft robotic arm to orient and control themselves automatically, to pick things up and interact with the world (Truby, 2020, para 6.) as well as, to progress onto artificial limbs that can dexterously control in the environment.

Even though the soft sensor is unable to capture precise movements, it still provides a steppingstone for machine learning in soft robotics control, enabling the experts to explore new models with improved sensitivity and control applications.  On the other hand, a soft robotic arm is more seemly than traditional rigid robot in the real-world application. Even though soft robot sensing technology has yet to be polished since it is a new field, the future aim for this technology is to develop a soft robot that is proficient in control application as well as capable of handling objects in the environment.

However, the one key problem for the preliminary sensors in the soft robotic arm is that they are unable to capture precise data. In the article, "Toward Perceptive...", (Wang, Totaro & Beccai, 2018), several computational approaches have been successfully implemented into the traditional robotics design such as the microelectromechanical system-based sensors. The infinite movements and the flexibility of the soft robotic arm will result in changes of the characteristics of a soft sensor attached on it. With their deformable body and high number of degree-of-freedoms, hence, the data will not be precise. (Case et al., Sensor configuration, 2018) The sensitivity of the sensor is affected due to the change in strain and pressure. In addition, the authors stated that it is impracticable to implement thousands of sensors in the soft robotic system due to problems in the components such as the electronics, sensor and space. Thus, the total number of sensors can be reduced by implementing the sensors at the best location for detecting multiple deformation.  

Other key elements must be taken into considerations in modelling of the sensor as well, such as the contact behaviour in between a sensorized soft robotic arm and an object that are difficult to anticipate as friction is involved, as there will change of pressure in the interaction of the object. Modelling of the sensor would be much more sophisticated, involving advanced method with high cost and the outcome might not correspond with physical systems. 

Secondly, the soft robot are not able to simulate full capability in their control application, thus an approach called "finite element methods (FEMs)" has been developed to introduce "real-time control algorithms"(Case et al., 2018, p. 31). Integrating the soft sensors with real-time loops are very beneficial, enabling more complicated tasks and more precise control of the soft robot. The sensory responses must be evaluated for a range of robotic tasks in function of the actuation mechanism, as well as depending on the scenario in which the robot is moving. 

Taking everything into account, soft robotic sensing is still its beginning phase and there are many challenges to overcome towards autonomous soft robots. These challenges could inspire new ideas for innovative solutions towards perceptive soft robots through scientific communities involving material science and biomedical engineering. Thus, a potential model may be feasible in the foreseeable future.

 

Reference List

Matheson, R (2020, 16 Feb).  Soft robotic arms uses flexible sensor to understand its position. Control Engineering. https://www.controleng.com/articles/soft-robotic-arm-uses-flexible-sensors-to-understand-its-position/

Wang, H., Totaro, M., & Beccai, L. (2018, 13 Jul). Toward perceptive soft robots: Progress and challenges. Advanced Science, 5(9), 1800541–n/a. https://doi.org/10.1002/advs.201800541

Zhang, Z., Dequidt, J., & Duriez, C. (2018). Vision-based sensing of external forces acting on soft robots using finite element method. IEEE Robotics and Automation Letters, 3(3), 1529–1536. https://doi.org/10.1109/LRA.2018.2800781