June 4 2020

ISSAI Data Scientist – Daulet Baimukashev’s – work has been presented at ICRA-2020

We are proud to announce that the research paper “Shear, Torsion and Pressure Tactile Sensor via Plastic Optofiber Guided Imaging” by Daulet Baimukashev (Data Scientist at ISSAI), Zhanat Kappassov and Prof. Huseyin Atakan Varol has been presented at one of the most prestigious scientific conferences – 2020 International Conference on Robotics and Automation (ICRA-2020).

This year, due to the COVID-19 pandemic, the conference is being held online. Therefore, all of the accepted papers are being presented by online means.

In their study researchers presented a new approach for sensing normal, shear, and torsional loads with an optical fiber-based tactile sensor which detects the displacements of a colored pattern. Similar to the human finger-pads, sensor is made of multiple non-identical elastic materials. Deformations of the elastic material are acquired using POFs and processed using machine learning.  

Research results can be used in many applications for surface exploration, object manipulation and force measurement. Traditional industrial robots are mostly pre-programmed and cannot interact with the environment. But using the tactile sensing, the control of the industrial robots can integrate the feedback from the environment for achieving more efficient and safe interactions.

“Most force/torque sensors available today have limitations that they cannot be used in hazardous situations as under magnetic field, underwater applications. However, we have shown that using the optical sensors which are robust to magnetic fields can be used to find the force/torque measurements. All the electronics and camera are placed away from the sensing part which distinguishes our sensor from the state-of-the-art methods and makes the sensor robust” – says ISSAI Data Scientist and one of the main authors Daulet Baimukashev.

Results of the study indicate that the sensor is capable of predicting localized normal force, shear and torsion with high accuracy. For example, contact location accuracy of the sensor is around 0.9-1.4 mm, while normal and shear force accuracy is 0.3-0.6 N. The accuracy for torque estimation is 5 mNm. Also, bandwidth of 4 Hz for contact feature estimation has been achieved. 

Research paper is being presented to the main audience of the conference, which includes a lot of fellow distinguishing scientists and researchers. The paper will be published on ICRA-2020’s conference proceedings for the future reference.

We congratulate our Data Scientist Daulet Baimukashev with this great professional achievement and wish him to work on more projects that would be presented at the world conferences! 

« of 2 »