The instrumented meeting room of the future will help meetings to be more efficient and productive. One of the basic components of the instrumented meeting room is the speech recording device, in most cases a microphone array. The two basic requirements for this microphone array are portability and cost-efficiency, neither of which are provided by current commercially available arrays. This will change in the near future thanks to the availability of new digital MEMS microphones. This dissertation reports on the first successful implementation of a digital MEMS microphone array. This digital MEMS microphone array was designed, implemented, tested and evaluated and successfully compared with an existing analogue microphone array using a state-of-the-art ASR system and adaptation algorithms. The newly built digital MEMS microphone array compares well with the analogue microphone array on the basis of the word error rate achieved in an automated speech recognition system and is highly portable and economical.