I've been thinking a little about this challenge and it raises a few question marks in my head,

Firstly - it seems that many of our existing tools - including the thousands of web2.0 aps are not designed to meet needs of human rights or advocacy organisations or indeed most people who want to make a change. Most are designed to fulfill an already saturated market. Even new mobile applications are basically gimmicks - look at http://code.google.com/android/adc_gallery/ for the big prize winners of the google android challenge. Amongst them are:  

• find a cab in new york
• finding the hottest party
• turn your workout into a race

Secondly -   it seems that most established applications are still very desktop/laptop computer centric with mobile as an afterthought. This is not suprising since they all grew up as desktop websites.

Thirdly - the successful mobile applications have all focused upon Chat and Messaging or ringtones/pictures or porn.

So the UC Berkley Challenge really cuts against the grain - Below a rough cut of an idea I have - If there is anything around like this please let me know (I want to use it).

If I get time and some support I may even put in an application, otherwise feel free to take elements that may be of interest: -

Working title : SMCMS : Rethinking Content Management from a community based and mobile device generated and viewed perspective.

Driving Concepts: -

• The Cellphone has changed communications and is reaching very high penetrations rate in all countries.
• Internet enabled cellphones are now more prevalent than Internet enabled computers in most developing countries.
• Most of these Internet enabled cellphones also have cameras and sound recording functionality.

This opens vast possibilities with respect to community made content and potentially for recording and publishing anything from human rights abuses to poor service delivery to collabrative entertainment.

To take advantage of this there should be an Internet based system which is designed to
      1) allow mobile content (text, audio, pics and video) to be uploaded from peoples phones to a central server
      2) convert this content into optimised sized/formated content for all other mobile devices together with standard web players. For example with video a 3gp version should be created which is optimised for any cellphone and a flv version should be created for standard computer web browsers.
      3) show this content in approriate sections in a format designed for cellphones
      4) allow an administrator to remove inappropriate content
      5) allow the content to be searched
      6) allow registered users to create mobile based 'microsites' using a combination of the uploaded content, text and links.
      7) allow links to these microsites to be SMSed or emailed to other people.
      8) allow microsites to be added to existing websites by simple php web scraping or by means of Wordpress, Joomla and Drupal modules.

With community based training - such a system will allow the generation of community based content and community news delivery which can address local issues more effectively than is currently possible.

If such a system is open source, simple to install on shared hosting and well publicised, large scale usage could be achieved quickly.

Challenges

• Data costs : Users must try to take pictures and videos at low resolutions to avoid high data charges.
• SMS costs   : The mobile links are most effectively distributed by sending sms links, however there are associated costs.
• Converting content : Converting video requires ffmpeg which is not available on most shared web hosting packages.
• Different Screen sizes : Most internet enabled Phones vary with regard to screen width : from 128 pixels to 320 pixels (iphone). These differences must be considered and coded for. Pictures should be resized to 4 different formats and displayed as required.  
• Abuse : If any system becomes popular it is often abused - There should be site controlled parameters for specific user based permissions. There should also be sensible Captas (non case sensitive) is required. Administrators must be able to moderator content before it is published.
• Difficulty of Login/Input : Entering a lot of text on a mobile device can be tricky. Therefore the amount of data that must be input should be configurable and passwords should be a combination of letters, numbers and symbols but should not be case sensitive.
• Quality of content : Because most users will be amateurs the quality of the content produced may be low. Target community based training will help.

From the above - The major technical challenge is the conversion of video and the major user factor is the cost related elements. Within a targeted programme (run by an NGO or other organisation) that uses SMCMS it may be approriate to give major contributors stippends. The video aspect may be temporarily covered by only allowing 3gp/3g2 video files   or may be covered by a dedicated server providing a web service for users of the application.