It is common in earthquake engineering research today to perform ad hoc checks on a theory by showing that it reasonably matches an insufficient experimental data set. It is not currently possible to collect and interactively explore enough information from experiments to fully characterize the complex phenomena involved. The number of instruments used in a centrifuge experiment is limited by the cost of instruments and signal conditioning. More restrictive, perhaps, is the complexity oflabor and bookkeeping involved in managing dozens of instruments. At UC Davis we are developing miniature data acquisition systems that can be mounted in the soil model to reduce the number and complexity of signal wires emanating from highly instrumented centrifuge models. The systems will be incorporated into a new high-performance data acquisition network.