RFID allows for data capture without human intervention, and for many tags to be sensed at once. This makes RFID data capture significantly more complex than reading bar codes or human data entry. Capturing meaningful data often requires the development or customization of software in order to orchestrate RFID devices with other sensors, human interaction and back-end systems. At the same time, the data-capture infrastructure must manage many RFID interrogators and keep them running, even if unattended. This session will explore how to architect a well-layered software system that separates these concerns, and illustrates typical RFID data-capture paradigms. The presenter will also introduce several data-capture standards, including LLRP, ALE, RM and DCI.