Should you build or buy RFID Middleware & Tracking?
RFID middleware sits between RFID readers and business systems, managing reader fleets, tag commissioning, deduplication, and the translation of raw reads into business events like received, picked, or shipped. Retailers, manufacturers, and logistics operators buy it to connect item-level RFID data to their ERP or warehouse systems.
For a single dock door or portal, a competent team can build the read-to-event pipeline on the Impinj IoT interface or LLRP, and some have. For an enterprise platform covering reader fleets, tag encoding, EPCIS exchange and ERP integration, no named independent team is documented running its own, and the vendors' licenses cost a fraction of the engineering it would take. Buy the platform, keep the business-event rules and integrations in your own hands, and reassess if open-source EPCIS 2.0 tooling matures.
Build it, buy it, or bridge?
When building makes sense
Building RFID middleware makes sense at a narrow scope: a single dock door, portal, or production line where a competent team can wire a reader's own IoT interface, publishing over MQTT or webhooks directly into a bounded read-to-event pipeline. Impinj's newer readers expose exactly that kind of interface, lowering the barrier for a defined use case. What has not been shown is a team building the enterprise version of this: a platform covering reader fleet management across many sites, tag commissioning at scale, GS1 EPCIS event exchange, and ERP or WMS integration. The open-source building blocks that once supported that ambition have gone quiet. Fosstrak's EPCIS repository, downloaded more than a thousand times and one of the most widely used open implementations, had its last commit years ago, and Impinj itself retired its own ItemSense middleware from support at the end of 2022. Build the narrow pipeline; buy the platform.
When buying makes sense
Buying earns its place for any RFID deployment beyond a single bounded pipeline. Reader fleet management, tag commissioning, GS1 EPCIS event exchange, and ERP or WMS connectors are exactly what platforms from Impinj, Zebra, Mojix, and Turck Vilant bundle together, and building that combination in-house means competing with vendors who have spent years on standards conformance most teams would rather not own. The case against building it yourself goes beyond cost: the open-source alternatives that might have supported a full build have stalled. Impinj discontinued support for its own ItemSense system software in December 2022, steering customers toward its partner ecosystem instead of a self-maintained stack, and the once-active Fosstrak project has not seen meaningful movement in years. Buy the platform for fleet management and standards conformance, and keep the business-event rules and system integrations as the part your own team owns.
The desk read
Build-versus-buy analysis for RFID Middleware & Tracking is being written. In the meantime, the framework that drives every B4 call is on the B4 Index page.
Frequently asked
What is RFID middleware?
Think of it as the translator between a reader that just sees tag IDs and a warehouse or ERP system that needs to know a specific event happened, like an item being received or shipped. Without that translation layer, raw RFID reads are just a stream of numbers with no business meaning attached.
When does building RFID middleware make sense?
For a single dock door, portal, or production line, where a team can wire a modern reader's own IoT interface directly into a bounded pipeline. It stops making sense at enterprise scale, where fleet management across sites, tag commissioning, and EPCIS conformance become a multi-engineer program rather than a weekend project.
When does buying make sense?
For any deployment spanning multiple sites, brands of reader hardware, or a real ERP or WMS integration. Vendors like Impinj, Zebra, Mojix, and Turck Vilant have already built fleet management and EPCIS conformance that a self-built platform would need years to match.
Why did Impinj retire ItemSense?
Impinj ended support for its ItemSense middleware in December 2022, and now builds reader intelligence directly onto its Speedway and R700 readers instead, leaving data processing and system management software to its partner network. Anyone still running ItemSense should plan a migration to a current platform.