Context-awareness and adaptation mechanisms is hard due to several issues. The context-awareness combine sensor fusion, adaptation and middlewares :) Furthermore, the context nowadays is everything, and the interaction with the user it the key purpose of most of the user-centric solutions (aka context-aware smart advisors), so you have to tech how to build the hybrid system that additionally will be intelligible and usable... 

Furthermore, adaptation could be rapid or gradual, and unless you have a lots of data that can reflect these types of changes in context, you have a problem. Additionally, such data may be sensitive -- location, user activity, user heart rate, applications they use, webpages they browse, etc. so sharing it with students may be illegal. 

For obtaining data, you may want to see: http://awareframework.com and http://www.kraken.me/#/home 

For context-awareness I have several tutorials that are easy to understand and use as teaching materials: http://glados.kis.agh.edu.pl/doku.php?id=pub:software:heartdroid:tutorials:start

They show how to use rules, basic statistics and temporal modalities, and machine learning if needed to build simple mobile context-aware systems in Java (also for Android) with HeaRTDroid  reasoning engine.

Thanks, Szymon.  This year, I introduced students to Google Awareness, and they also have implemented some Android apps with a middleware we developed for context management (LoCCAM). However, I am still in search of the best organization of the Ubiquitous Computing course. I believe there is an absence of tools that support theoretical and practical classes. One of my master's students is developing a tool for this purpose. I'll check the documents you sent me. Thank you very much.

Windson, If the software your student work on will be open for academic use, I would be very interested in it.  I looked over LoCCAM, but it seems that all the documentation is in Portugese :) I have never used Google Awarenes -- we used Aware Framework instead (I have linked it to you), as it itself is using Google API to some extent, and provides more context via plugins.

I took a look at Aware; it will fit very well in my course. Thanks. Regarding the tool of my student (Joelma Peixoto), it will be a free tool for educational purposes, and this time with a good description/version in English ;).

When she has done the work, I will send you a link to evaluate the tool. Regards!

Ps: Could you please answer a short questionnaire for us? Joelma has elaborated it for her requirement elicitation process.

https://docs.google.com/forms/d/e/1FAIpQLSekCswgCeGlMEkRM47e0fj8SjrflDWs4ph6GTNYQWpAfd6ecQ/viewform

Ps2: I have attached to English papers about LoCCAM

Article LOCCAM - Loosely coupled context acquisition middleware

Conference Paper CRITiCAL: a ConfiguRation Tool for Context Aware and mobiLe ...