Accessing Remote Laboratories from Mobile Devices

Introduction

A Remote Laboratory is a software and hardware system that enables students to use real experiments physically located in a university. This way, students can access real experiments 24 hours a day, 7 days a week, even including holidays, from anywhere with access to the Internet. Given that the experiments are real, the Remote Laboratories have the opportunity to probe this fact to the students (i.e., with a webcam), so the students don't lose the feeling that they are doing exactly what they would do in a hands-on-lab session.

As an example, WebLab-Deusto, a Remote Laboratory developed in the University of Deusto, counts with different experiments. In the FPGA experiment, the student works with the Xilinx IDE at home, develops a program for a FPGA device. Once the student has finished programming it, he can connect to the Remote Laboratory, send the program, which is programmed in a real FPGA, and finally interact with it with the panel of Figure 1. During a limited amount of time, the user will be able to modify some inputs of the device and see through a webcam the produced outputs. The students are not limited in how many times they can use it, but the amount of time for each use is limited so the device can be reused by other students.

Figure 1.

Sample screen of WebLab-Deusto

Remote Laboratories also provide an efficient performance of the devices for the university. Many experiments do not require a continuous connection to the device. For instance, in the FPGA experiment described above, the device is only used by the student during the small amount of time in which the device is programmed and tested. This amount of time is enough for testing if the code was correct or what failed. After this time, the user can go back to the IDE and try to correct the problems, and test again connecting to the experiment. If somebody else is using the device, the latter user will enter in a queue of users and will stay there until all the users that entered before him have finished. If there are many users, the university can put several copies of the same experiment, so the students will automatically use different experiments and the mean time in queue will decrease. Due to this transparent and efficient sharing among different remote students, the Remote Laboratory might achieve higher levels of throughput per device than traditional laboratories.

As an example of a real scenario, during the course 2008-2009 in the University of Deusto the CPLD experiment had 74 users, and the FPGA experiment had 23 users. Figure 2 represents the occurrences of each position in a queue for the two experiments during that course. This is, if a student entered in a queue in position 5, Figure 2 represents one occurrence for position 5, and no occurrence for lower positions, even if he stays in the queue and achieves those lower positions. It also represents only the positions in the queue, from 0, which means that the student will use an experiment as soon as a device becomes free, to 19, which was the maximum achieved position. Therefore, the figure does not represent the number of students that directly used the experiment without being in queue, which was the majority. Furthermore, during this course there was a single FPGA device and a single CPLD device, so since WebLab-Deusto balances the load of users among the available devices, it is predictable that the distribution would have varied considerably if more devices had been placed. However, it is still remarkable that given that the maximum session time was 200 seconds, a single device enabled students to wait less than 10 minutes 87% of all cases where the user entered in a queue, getting a higher percent if we counted the users that didn't even enter in the queue.

Figure 2.

Relation between positions in a queue and how many occurrences of those positions in CPLD and FPGA experiments of WebLab-Deusto during the course 2008-2009