Molo is involved in a number of research projects which are not directly related to our product range. These projects are generally aimed at exploring and improving the use of speech technology in a developing-world context, and are performed in collaboration with the University of Pretoria and the Council for Scientific and Industrial Research (CSIR). A few of these projects are summarized below.

Open Source software has created a revolution in various domains of the computer industry, and open-source components have been developed for various telephony and speech functions. This is very exciting for developing-world applications, since proprietary software systems are generally not affordable for such applications.

With our partners we are developing a turnkey solution that is easy to use, comparable to the state-of-the-art systems in performance, and based on open-source components. We are also collaborating with various institutions to develop public-domain resources that will make it possible to apply such a solution to several South African languages.

In many developing countries (such as South Africa) the availability of mobile telephones has created entirely new opportunities for providing information to technologically underserved communities. Internet penetration tends to be low in such communities. If one can provide useful information to members of such communities over the telephone, a whole range of hurdles can be side-stepped. However, relatively little is known about the design principles that will optimise the usability of such interfaces. For example, when should speech recognition be used, and when is input using keypresses preferable? How should prompts be phrased? These questions can only be addressed by careful experimentation involving diverse user populations. With the CSIR, we are developing a research programme that aims to provide systematic answers, which will hopefully be useful to all developers of telephony systems aimed at technologically inexperienced users.