The Responsphere testbed and infrastructure consists of three components (hardware, software, and disaster response exercises) as well as three stakeholder groups (industrial partners, government partners, and other academic/researcher partners). The hardware portion of the infrastructure is truly a unique testbed for disaster science and supports research ranging from the social sciences to algorithms. In order to support this research, the Responsphere team has created a sensorized and instrumented sentient space that encompasses approximately 1/8 of the UCI campus (Emergency Response Zone 4) with our eventual goal to instrument fully 1/3 of campus. Our instrumentation ranges from high-end computational systems such as a 22-node Beowulf cluster (donated by IBM), two 8-processor computing systems supporting Solaris and many virtual machines, to very low level sensors such as load cells that measure pressure. Other types of sensors created by the team include the EvacPack (human-as-a-sensor with instrumentation such as Gieger counters, temperature, light levels, magnetometers, acoustic sensing, gas sensing, and accelerometers), to robotic (autonomous and semi-autonomous) sensing platforms, large-scale smart-building sensors with optical, motion, acceleration, "people counters", that tie into the power-grid information which includes utilizing the power lines as network backhauls, to a completely pervasive outdoor 802.11 network. Also within this outdoor environment are optical and acoustic sensors that are deployed in ruggedized enclosures. Finally, Responsphere houses possibly the world's largest collection of disaster response data sets. We have close to 4TB of data available to disaster scientists (with the proper IRB approvals) ranging from sensed data during drills to 911 calls from all over the country.