{"id":532,"date":"2021-12-24T11:53:14","date_gmt":"2021-12-24T11:53:14","guid":{"rendered":"https:\/\/MohamadSaada.com\/?p=532"},"modified":"2021-12-24T21:52:53","modified_gmt":"2021-12-24T21:52:53","slug":"yoban","status":"publish","type":"post","link":"https:\/\/MohamadSaada.com\/?p=532","title":{"rendered":"Yoban"},"content":{"rendered":"\n
\n

I have worked for two years on the \u201cYOBAN<\/a>\u201d project which was an Innovate UK<\/a>\/Newton Fund<\/a> project. The objective of this project was to develop a pre-production level prototype of an assistive robot, called \u201cYOBAN\u201d, which will provide walking and sitting assistance and useful functions similar to companionship aimed at improving the elderly user\u2019s quality of life.<\/p>\n<\/div><\/div>\n\n\n\n

\n

The technical objectives can be broken down into three areas; those challenges relating to the robot\u2019s electro-mechanical, computing and control systems, those challenges related to the interfacing to the human user and those challenges related to Internet services. The prototype robot would assist the user in making short distance travel outdoors. A home version will act as the user companion, it will be able to respond to hand gestures to make calls to a carer\/family member or to play games or play internet audio content. The work that I have completed on this project involved two phases. Phase one focused on Robot control and this involved writing low-level code to control the (Atmel) motor controller boards, as well as writing higher level code for the balancing Robot as well as for the electrical steering robot. Phase two involved developing deep learning models for three distinct purposes.<\/p>\n<\/div><\/div>\n\n\n\n

\n
  1. Detecting road surface type using IMU data and an LSTM network model.<\/li>
  2. Hand gesture recognition using a custom-built deep learning detection model using depth and colour data from the Intel Realsense Camera.<\/li>
  3. Facial recognition model using Siamese neural networks for recognising the owner of the robot.<\/li><\/ol>\n<\/div><\/div>\n\n\n\n
    \n

    The YOBAN project was a successful project that involved a partnership between four different entities, one of them is Loughborough University and the others are a robotics company in Southampton and a highly regarded research institute in China and a manufacturer in China. At the end of the project, all teams that are involved in the project travelled to Shenzhen China, where the successful trials for the project were conducted in a rehabilitation centre. The outcome of the project was a great success and the UK side obtained two patents for some of the work that has been done.<\/p>\n<\/div><\/div>\n\n\n\n

    \n

    China Trip<\/h1>\n<\/div><\/div>\n\n\n\n
    \n

    In July 2019 near the end of the project we have travelled to Shenzhen, China. The purpose of this trip was to conduct trials for two of the three main types of the robot, The “Trolley YOBAN” and the “Companion YOBAN”. Prof. Meng, Christos and I travelled from the Loughborough side. Dennis and Joel travelled from the Motion Robotics Side. The robots were shipped before hand by Motion Robotics.<\/p>\n<\/div><\/div>\n\n\n\n