Yazar "Ayyildiz, M." seçeneğine göre listele
Listeleniyor 1 - 4 / 4
Sayfa Başına Sonuç
Sıralama seçenekleri
Öğe A Low-Cost 3D-Printed Soft Pneumatic Actuator(Institute of Electrical and Electronics Engineers Inc., 2021) Tüysüz, B.; Güney, E.; Serpiciler, D.; Ayyildiz, M.Soft actuators have recently gained popularity due to their advantages over rigid actuators, such as contact compliance, endurance against mechanical failure, and better compatibility for human-machine interaction. They enable robots to handle fragile objects safely and interact with their environments effectively. However, the production of cost-effective soft actuators is challenging due to the complex manufacturing procedures of complex actuator architecture. In this study, a custom-made 3D printer has been developed for printing flexible filaments to produce pneumatic silicone actuators. First, soft pneumatic actuators with programmable bending motions were designed, and this capability was characterized through FEM simulations using ANSYS Mechanical. Then, these actuators were 3D printed and tested under the same conditions used in simulations to validate their characteristics. The results showed that the simulations and actual experiments were in good agreement. In the future, we are planning to produce soft pneumatic robotic grippers by combining three of the proposed soft actuators. © 2021 Chamber of Turkish Electrical Engineers.Öğe Development and Characterization of a Wearable Ring Providing Haptic Feedback(Institute of Electrical and Electronics Engineers Inc., 2022) Mayo, F.P.L.P.; Canbek, K.O.; Ayyildiz, M.Comfort and efficiency are the essential features of any wearable device. A variety of devices use vibration to provide haptic feedback to the user. In this study, a novel haptic flexible ring equipped with four actuators is developed to enhance the user experience for tactile interfaces. The proposed device provides subcutaneous clues at the proximal interphalangeal joint of the index finger to mimic the forces and moments applied to this joint. Thresholding experiments are conducted using a total of nine subjects to evaluate the effects of the vibration amplitude, thickness, and orientation of 1D lines on human perception. The results showed that as the line's thickness and amplitude increase, the participants perform better at identifying the shapes. The minimum thickness that the human participants can detect is 2 mm at the lowest possible vibration level with a single motor. Furthermore, the second vibration amplitude and vertical orientation lead to higher recognition rates and lower recognition times. In the future, we aim to perform further characterization experiments and develop virtual reality applications to test the proposed ring. © 2022 IEEE.Öğe Handwritten Digit Recognition using Spiking Neural Networks(Institute of Electrical and Electronics Engineers Inc., 2022) Ozcan, O.; Oniz, Y.; Ayyildiz, M.In this work, both simulated and experimental studies have been carried out using Spiking Neural Networks (SNNs) on the handwritten digit recognition problem. The design of SNN is performed using Spike Response Model (SRM). A gradient based algorithm is applied for the learning of SNN. For the simulations, the proposed algorithms have been applied on the MNIST data set. To provide a basis for comparison, the same studies have been also performed for an equivalent non-spiking artificial neural network (ANN) structure. Following the simulated studies, experiments utilizing a multi touch IR frame have been carried out for both network structures. Two different approaches have been adopted to evaluate the performance of the proposed network models: In the first one, the training of the network structures has been accomplished on the MNIST data set, whereas the data acquired from the experimental setup have been used in the testing phase. Next, the data obtained from the experimental setup have been employed in both training and testing of the neural networks. The results of the simulations and experimental studies reveal that the SNN outperforms the conventional ANN. © 2022 IEEE.Öğe The Effect of Book Preconditioning on Page-Turning Success Rate during Automated Book Digitization(Maik Nauka/Interperiodica/Springer, 2022) Sinmaz, E. K.; Kocasecer, M.; Ayyildiz, M.The growing popularity of e-books and the adoption of e-book reader devices have increased the need for book digitization. Traditional flatbed scanners are easily accessible and widely available; however, manual scanning of books is tiring, tedious, and time-consuming when the book has many pages. Developing automated book scanning instruments at a low cost can allow more people to efficiently access and execute the book digitization. In this study, we propose a cost-effective book scanner with an automated page-flipping function. Our prototype uses a motorized moving shuttle with a vacuum gripper to hold the page and a fan to ensure the turning of the page in the correct direction and number of times. A Raspberry Pi 4 takes pictures of the left and right sides of the book using two digital cameras. When the picture capturing process is completed, it combines pictures into text-based digital formats such as PDF using OCR technology. In our experiments, we scanned five different books having different page numbers (120-700) with various paper characteristics such as paper size (A4 and A5), paper type (coated vs. uncoated), and paperweight (60-250 g/m(2)). The results showed that we obtained less than 0.9% error in book scanning when the paperweight was between 60-120 g/m(2). We achieved the most successful scanning with uncoated paper type, A4, A5 paper size, and 115-125 g/m(2) paperweight. We observed that the unsuccessful page scan rating decreased as the number of scanning repetitions increased. The results also showed that it was challenging to turn the pages of new books without encountering problems unless pages were preconditioned by turning. We concluded that three iterations of preconditioning are necessary for new books to reduce errors in the page-turning process.
