Application to measure sensor data and various parameters - ideas and questions

[Nadina Sovaiala]

Hello,

My name is Nadina and I am a second year Master Student at VU Amsterdam, Computer Science, track High Performance and Distributed Computing. My current research topic is distributed smartphone computing,  in particular context aware applications, both local and on remote devices. I also have industrial experience in mobile development.

I am interested in your application for measuring sensor data and various parameters, field on which I am currently working for my thesis. I consider that my knowledge gained so far would suit best  this topic. I have several questions and suggestions. In case I get a clear idea of what this application should do and if my suggestions appeal to you, I will formally apply for this project at GSoC. 

1.Should the application function offline (Internet connection is down), meaning that the sensors should be stored locally and all the processing will be done locally?

2."the value of sensor data / computed parameter could be plotted against time." -> How long should the interval be ? If we are talking of a big interval, then we would need specialized storage for large amounts of data.

3.Should the focus of the app be on the way the retrieved data is represented ?

4.Should the app only consider device sensors or can it retrieve information from external sources (I think the second would be more appropriate if we want the analysis to be more accurate) ?

A first suggestion would be to use CommonSense for storing big amounts of data from the sensors. The API provided offers many other functionalities, like grouping, sharing or data analysis. I am already familiar with this platform, as I used it in my research projects. 

Or the app could make smart decisions based on some heuristics upon where to store the data (on cloud it it has wifi connection, else locally).

 

Thank you,

Nadina Sovaiala

[Pankaj Nathani]

Hi Nadina,

 

My name is Nadina and I am a second year Master Student at VU Amsterdam, Computer Science, track High Performance and Distributed Computing. My current research topic is distributed smartphone computing,  in particular context aware applications, both local and on remote devices. I also have industrial experience in mobile development.

I am interested in your application for measuring sensor data and various parameters, field on which I am currently working for my thesis. I consider that my knowledge gained so far would suit best  this topic.

Great, if you are working on a thesis for measuring sensor related data.. you must already have a more granular idea of their functioning.

 

I have several questions and suggestions. In case I get a clear idea of what this application should do and if my suggestions appeal to you, I will formally apply for this project at GSoC. 

1.Should the application function offline (Internet connection is down), meaning that the sensors should be stored locally and all the processing will be done locally?

Yes, as of now storing the recorded data is local. You may provide options to export the recorded data to a file which can then be tranferred to PC or another device.

 

2."the value of sensor data / computed parameter could be plotted against time." -> How long should the interval be ? If we are talking of a big interval, then we would need specialized storage for large amounts of data.

Yes, most of the parameters would be plotted against time. Typically 1 sec would be the interval for real time. However, you can have this value to be customisable by the user.

 

3.Should the focus of the app be on the way the retrieved data is represented ?

Yes, visualisations are important. We'll like your proposal more if you are more creative on the visualisations.

 

4.Should the app only consider device sensors or can it retrieve information from external sources (I think the second would be more appropriate if we want the analysis to be more accurate) ?

What do you mean by external sensors? You mean like a bluetooth sensor? We do not need the info from external sensors for this version of the application / for the scope of GSoC. But may be in the future versions we can think of adding.

 

A first suggestion would be to use CommonSense for storing big amounts of data from the sensors. The API provided offers many other functionalities, like grouping, sharing or data analysis. I am already familiar with this platform, as I used it in my research projects. 

Would this involve storing the data on the cloud? If yes, we may not need to use it since we are only doing local storage. However mention it in your proposal and it's advantages. We can consider this suggestion.

 

Regards,

Pankaj.