Teleoperated by an Open Source Android Application

Teleoperated by an Open Source Android Application

Humanoid RobotThe Humanoid Robot we opted was NAO [14] fromAldebaran Robotics. Aldebaran Robotics is a French-basedcompany, gain popularity in 2004 with a new novel projectknown as NAO. NAO is running OpenNAO OS which is aGNU/Linux distribution based on Gentoo. It supports bothwired IEEE 802.3 and wireless IEEE 802.11 a/b/g/n connec-tion. With a lithium battery inside, has autonomy up to 90minutes (active use). NAO is equipped with two ultrasonicsensors (or sonars) which allow it to estimate the distanceto obstacles in its environment. Also, two identical videocameras are located in the forehead which they provide anup to 1280×960 resolution at 30 frames per second. As aresult, NAO can be a highly effective platform for teleroboticswhen its user tries to start an interaction with people or objectsnearby the robot.A user of our system can manipulate the robot using anAndroid smartphone by observing the surroundings of it froma video feed played on the smartphone display. We createsoftware that is executed on NAO in form of Python and Bashshell scripts files that are necessary for the complete set ofour programmed activities. In our embedded software, we usean ever ended loop that polls the bridge web server everyone second to receive the next command for execution. Afterthe successful execution of the specified command, the robotacknowledges it to the web server in order to be removedfrom the execution queue of the web server. This feedbackmechanism is crucial for the correct execution of the activitiesand ensures that there are not lost events. The communicationbetween the robot and the bridge server is performed by adecimal arithmetic protocol that we have developed in-house.DAndroid ApplicationTo control the humanoid robot we have developed anandroid application. The user has full control of the robot likemove forward, turn left etc. Except of these simple commands,our system has the capacity of speech to text, text to speech,video and audio streaming. With these new features, we havecreated a new way of communication between a blind user anda speech-impaired user. The humanoid robot is the key to thatcommunication by speaking out loud the messages users wantto tell or inputted via text respectively. Except of this new wayof communication, our system can be used for safety purposes.For instance, a father can watch what his child is doing whenhe is not at home.In the first screen of the application there is a button whichis used for the connection between the application and theserver. Also, the user can change the configuration settings bypressing the settings button.The main screen of the application is shown on Figure2. It consists of multiple action buttons which are used tocontrol the NAO. Every time the user presses a functionalitythe Android device sends an HTTP POST request to the bridgeserver and the server appends this to an activity queue. Everytime the humanoid robot polls for a new activity the web servertransmits the activity on the top of the queue. Also, on top ofthe screen, is being displayed video or pictures that are beingtaken by NAO.Our application also has two unique features. Text to speechand speech to text capabilities. In Figure 3 our humanoid robotis giving a voice to people with speech disabilities.The Android application was created in Android Studio.The main programming languages that were used in orderto be created successfully are Java and XML. The minimumSDK for the android application is API 16: Android 4.1(Jelly Bean). Some of the most important libraries we usedare Apache library which was used for the HTTP requestsbetween the android device and the web server, Picasso librarywhich was used to load images from web URLs, AudioFormat,AudioRecord, MediaRecorder that were utilized for the WAVrecording. CodeShoppy

 Teleoperated by an Open Source AndroidApplication

To evaluate the efficiency of our implementation, we arereferring a number of metrics that we have gathered. Specifi-cally, we used our application from the Polytechnic Universityof Catalonia in Barcelona, Spain, in order to remotely controlour Humanoid Robot that was located in Kozani, Greece. Theaverage time it took until the request was made by the mobileand get a response from the server was 0.233 sec. Also, Theaverage time it took for the Audio Streaming was 0.369 secand 0.686 for the Video Streaming. This was measured usinga performance metric function of our android application fromthe time it posted the command to be executed until the time itreceived the successful reply message. Except from Barcelona,we also tested our application from University of Thessaly inVolos, Greece, in order to compare the results. The averagetime it took until the request is made by the mobile and get aresponse from the server was 0.16 sec. Also, the average timeit took for Audio Streaming was 0.347 sec and 0.505 for theVideo Streaming. As we notice, the small differences betweenthe delays from Spain and Greece mean that the application isable to run from all around the world without any considerabledelays.To decide which User Interface (UI) to implement on ourAndroid Application, we surveyed more than 50 people toask them about the user experience that they had over threedifferent User Interfaces that we have created. The age rangeswere: under 18, 18-25, 26-40, 41-55 and 55+. We wanted theapplication to be easy to use from all of the above ages. Theresults of the statistical analysis were that the majority of theusers preferred the minimalist version of the proposed UI’s.Taking into account this survey we ended up with the aboveapplicationInternet of Robotic Things allows robots or robotic systemsto connect and share all under a hood of a sophisticatedarchitectural framework. Our paper proposes an open sourceandroid application for controlling a humanoid robot. Ourapplication has accomplished so much more than any of itspredecessors. The user can view what the robot is seeingthrough the Android application and can also move it. Ad-ditionally, our system can deliver real-time audio through ourapplication. With the use of text to speech and speech to text,our humanoid robot is giving a voice to people with speechdisabilities. In the future, we plan to enhance the system byinserting more functionalities such as face recognition and AI

https://codeshoppy.com/php-projects-titles-topics.html