Design of a Mobile Robot to Work in Hospitals and Trajectory Planning Using Proposed Neural Networks Predictors


YILDIRIM Ş., SAVAŞ S.

International Conference on Reliable Systems Engineering, ICoRSE 2021, Bucharest, Romanya, 9 - 10 Eylül 2021, cilt.305, ss.32-45, (Tam Metin Bildiri) identifier identifier

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Cilt numarası: 305
  • Doi Numarası: 10.1007/978-3-030-83368-8_4
  • Basıldığı Şehir: Bucharest
  • Basıldığı Ülke: Romanya
  • Sayfa Sayıları: ss.32-45
  • Anahtar Kelimeler: Mobile hospital robot, Omni-drive, Controller design, Trajectory tracking, Neural networks
  • Kayseri Üniversitesi Adresli: Hayır

Özet

© 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.Considering the intense and tiring working conditions in hospitals, healthcare personnel’s performance decreases during prolonged working times, and patients are directly affected by this decrease in performance. This study aims to design and implement a mobile robot that can help healthcare professionals improve the healthcare industry conditions. In this context, the focus is on the mobile robot performing two main tasks. The first task is dispensing medication to patients with an eight-chamber mechanical feeding unit. Thus, patients can take only their medicines from the defined reservoir by selecting their names or photos on the touch screen. The second task is to interact with patients to give moral support with phrases such as “good morning”, “you look great today”. Also, drug delivery activity is recorded in a database, and the health status of the patients can be kept under surveillance with the camera on the mobile robot. The designed mobile robot goes to the patient rooms with magnetic strip tracking. For this purpose, a controller is designed for the omni-drive robot using MATLAB, and its performance is simulated. Also, the control velocities that enable tracking the trajectories are taught to artificial neural networks (ANN), and the requirement magnetic strip for trajectory tracking is eliminated. In this direction, two artificial neural networks are compared in terms of their learning performance.