FBG-based sensing systems for robotic manipulators to monitor interactions in unstructured environments
Name: VITORINO BIAZI NETO
Publication date: 14/03/2022
Advisor:
Name |
Role |
|---|---|
| ANSELMO FRIZERA NETO | Co-advisor * |
| ARNALDO GOMES LEAL JÚNIOR | Advisor * |
Examining board:
| Name |
Role |
|---|---|
| ANSELMO FRIZERA NETO | Co advisor * |
| ARNALDO GOMES LEAL JÚNIOR | Advisor * |
| MARIA JOSE PONTES | Internal Examiner * |
Summary: The emergence of Industry 4.0 has brought new concepts to the factories that optimize and improve conventional processes. These technologies have brought assignments to the industrial robots that allow them to perform tasks faster and more precisely. The
improvement of the robots proprioception capacity and tactile sensitivity through the use of sensors is a useful approach to achieve those goals. Optical fibers are a viable technology to be used as sensors in robotic devices because they are electrically passive and present electromagnetic immunity. This dissertation proposes two Fiber Bragg Grating (FBG) based sensing systems to monitor robotic manipulators during their operation. The first system are instrumented claws which work based on the strain rosette strategy to measure the critical stress-strain state on its surface while an object is being picked up. The other one consists of a group of smart textiles installed on the robots body to detect interactions with the environment. Three FBG sensors are positioned according to a stain rosette configuration within a slot in two 3D printed claws. The equations based on elasticity theory to obtain the stress-strain state from the sensors data are developed and a methodology using an auxiliary temperature sensor to prevent cross-sensitivity and to avoid measurement errors is proposed. The characterization tests show a highly linear behavior in terms of sensitivity to strain (R² = 0.996 for the linear regression of the means) and to temperature variation (lower R² between FBGs equal to 0.9868). From the compression tests, it is possible to conclude that the dynamic range of the sensors is adequate to the load capacity of the claws. Grip tests with different objects were performed on the manipulator to verify the systems feasibility in real-time applications. The tactile system consists of a group of optical fibers with multiplexed FBGs embedded in silicone rubber. The optical fibers with the sensors are positioned between layers of polyethylene foam and cotton fabric. A mathematical model is proposed to determine the maximum distance between FBGs on the system to achieve the systems best spatial resolution, however this is not followed due to cost and equipment limitations. After the manufacturing process, temperature and force characterization presents almost all values of R² on the linear regression superior to 0.94. Individual analysis are performed for the sensors which presented low coefficient of determination. Finally, on the experimental validation of the system, it can be observed that it can provide the position on robots body, the amplitude in terms of force and instant of time which a external impact occurred.
