Abstract:

Robot models have been used for a long time as case studies for exemplifying general concepts of analysis and design in many fields like systems engineering, electrical engineering, control engineering, computer engineering and other. We are facing in the last years an unprecedented development of robotics and of its contribution in the study of the aforementioned fields. Robot models are now easy to be obtained in universities because of the affordable manufacture using 3D printing and because of the availability of effective and inexpensive devices for interfacing and control.

We describe in this paper the steps performed by students from the Faculty of Engineering in Foreign Languages of the University POLITEHNICA of Bucharest for their undergraduate theses in order to create an anthropomorphic robotic hand and to control it using the available technologies. The construction process is much simpler than the traditional approaches and can be performed by 3D printing of the hand pieces, assemble them and connect servos for fingers movement. The control signals are sent by an Arduino board and there are tried several methods to collect the fingers data from the human operator. There are described and compared the three methods used to command the hand. The first one is direct - a glove is mounted with flex sensors that changes their resistance value when bent. The flex sensors are connected to the Arduino, which commands each servo to produce a similar bending to the one present in the glove. The second and the third methods use a PC to get the data and the Arduino board takes just the command signals from the PC and converts them to control signals for the servos. The second method uses the DGTech DG5 VHand 2.0 data glove with flex sensors and Bluetooth communication to take the fingers bending, while the last one uses a Leap Motion device that acquires wirelessly the hand posture, including the fingers bending. Based on several trials that were performed upon the prototype, using both the sensor gloves and the Leap Motion, several errors were identified and corrected, insuring valuable lessons of control engineering to the students, in a learning by doing approach. In the end, there are depicted future improvements to the system, tied to the particular control methods.

Keywords:

Robot hand control, Data glove, Leap Motion.