• Installation and reader management

    • Provided the ability to specify a list of reader addresses for provisioning rather than scanning the entire network for readers
    • Automatically generate SSL certificates to secure southbound API communications at installation time, and whenever certificates are missing when the proxy container starts.
  • Management and configuration

    • Allow end users to enable and disable secure IMC, AMQP, and northbound REST API communications using their own SSL certificates
    • Added the ability to export & import ItemSense configuration records within the ItemSense Manage Console
    • ItemSense now provides an NTP server container and a new REST API method that will enable / disable reader synchronization to the NTP server
    • Improved log file rotation and size policies for better disk usage handling
    • Improved database pruning strategies to work down to a minute resolution rather than day resolution for better disk usage handling
    • Added functionality to update the reader name via a new API endpoint
    • Removed the ability to change the reader type in reader definitions once they have been created
    • Improvements to facilities, floors, and zones:
      • Reader zones must now have the same name as the reader definition. If a reader zone is not supplied when created, it will be inserted automatically
      • Fixed reader definitions so that the facility or floor can be changed without having to change antenna or reader zones
      • When using the IMC, it is required to specify the information when creating a reader definition
      • When using the API, it is not required to provide reader placement and floor information when calling the reader definition create API method. If none is specified, the default floor for the reader's facility will be used
    • Added functionality to download remote reader logs generated by Octane firmware and the reader agent using either the ItemSense Management Console or API endpoint
    • Added functionality to reboot any managed reader using either the ItemSense Management Console or API endpoint
    • API Improvement - If the reader zone is specified when creating or updating a reader definition, it must match the reader name
  • Data and Performance

    • Ability to run multiple inventory & location jobs in one facility, for overlapping tag populations
    • Added support for firmware-based power sweep when running in inventory mode
    • Added support for two tag filters with an "AND" or "OR" combination for inventory and location reader configurations
    • Allow filtering by multiple jobs when querying Items & ItemHistory table records.
  • Threshold algorithm:

    • Multiple accuracy improvements to threshold algorithm
    • Support for threshold lane beams to improve OverHead Offset confidence


  • Installation and reader management

    • ItemSense setup script is now a self extracting .run file rather than a .bz2 archive. See the Installation Guide for details
    • ItemSense is now configured to run as five docker containers ((ItemSense, IMC, NGINX, MariaDb, and RabbitMQ) rather than two, in order to use the officially maintained images of standard components, as well as provide a more modular, maintainable architecture
    • ItemSense uses host networking rather than docker networking. Refer to the ItemSense Requirements document for the required ports
    • The protocol between ItemSense and its managed readers is now robust to changes in reader IP address
  • Management and configuration

    • Supported regions now include China and Japan
    • Default ETSI region threshold reader configurations and recipes now included
    • New Factory Reset endpoint
    • New Item Reset endpoint
    • Added ItemSense Management Console log to Log Export
    • The ItemSense version file and build information is now included in log export
    • The JobRunner role has read access on all endpoints
    • Thresholds can only be deleted if all of their referenced readers are idle
    • Reader groups cannot be specified on a threshold job
    • A reader placement can be updated when the reader's floor is null
  • Data and performance

    • Added a deliveryMode AMQP message persistence flag on the Queue Creation endpoints
    • Web server logs now rotate daily or when they become too large
    • Health events now survive database pruning
  • Threshold monitoring and tuning

    • Introduction of a real-time threshold monitoring page in the ItemSense Management Console
    • Upgraded threshold algorithm version to match the latest tuning tool
    • Use of tag status to aid threshold detection
    • Use of overall ratio to aid threshold detection


2017r1 build 850.5

  • Minor improvements and bug fixes
  • Optional Job parameter reportToFileEnabled has been marked as obsolete and no longer has any effect

2017r1 build 830

  • Installation and reader management

    • Installation time decreased by several minutes
    • Installation directory displayed during installation
    • Improved handling of readers with skewed or misconfigured system clocks during registration
    • Fixed provisioning limitations that processed one reader at a time when using non-default reader passwords
    • Prevent registering readers with an ItemSense instance with a host address of localhost, to avoid pairing failure
    • Allow more than 10 readers to be upgraded at one time
  • Configuration

    • Improved default antenna configuration for overhead-mounted readers in threshold setups
    • Allow floor to be set in the placement configuration of all readers, not just xArrays
    • Remove Tag Expiry as an option for Single Target B-to-A inventory jobs since tags only answer once in that mode
    • Enhanced EPC Gen2 filters to work with odd as well as even mask lengths
    • Added output of "soft" transition events for helping tune the parameters used in Threshold algorithm
  • Data and performance

    • Handle large tag population with improved data accuracy
    • Enhanced job processing to ignore results from previous jobs, so that message queues provide results that match ItemHistory tables
    • Improved job reliability concerning transient network failures, automatically terminating stale jobs, restarting interrupted RFID jobs, and ensuring resumption of data after re-connection
    • Fixed failing parallel job starts (when starting multiple jobs at the same time through the API)
    • Fixed reader communication failures (413 errors) when sending too much data (> 4MB) in location mode
    • Fixed 'null' floor designation in Message Queue events
    • Fixed 'Agent Initialization Error' in Location jobs when Z axis is not specified for reader
    • Fixed tag processing null pointer exceptions when the same tag is processed by a Location Job after an Inventory job
    • Better tolerance of "out of disk space" file-system conditions
  • Search and export

    • Fixed ItemHistory CSV export for large datasets
    • Fixed Items CSV export when specifying a Facility without a Zone
    • Fixed Zone Transition Only ItemHistory search

2017r1 build 697

  • Threshold crossing detection for Dock Doors, with the following reader configurations:
    • Side by Side - Readers flanking either side of the threshold, facing inwards
    • Overhead - One or more readers directly overhead the door
    • Overhead Offset - one or more readers set overheader but in between thresholds
  • Ability to run multiple jobs in one facility, for non-overlapping reader fields of view
  • Ability to run jobs on one or more pre-defined reader groups
  • Ability to download and save ItemSense configuration
  • Ability to download log files
  • Support for Antenna Hub hardware component connected to each antenna port of a Speedway reader
  • Support for new Search Modes:
    • SINGLE_TARGET_BTOA: Single Target Reset Inventory
    • DUAL_TARGET_WITH_BTOASELECT: Dual Target Inventory with Reset
  • New API endpoints
  • New AMQP queue
  • Changes to existing API endpoints
    • Addition of groups field to the Reader Definition API
    • Addition of features field to the Reader Definition API
    • Addition of THRESHOLD value for operation field of the Reader Configuration API
    • Removal of deprecated NORMAL value for the operation field of the Reader Configuration API
    • Addition of THRESHOLD Recipe type
    • Addition of name and readerGroups to Job API
    • Modification of the default value of the graceful parameter on the Start Job endpoint from false to true
    • Addition of jobId to Items APIs and Queue data
    • Date and time formats all use the same ISO-8601 format. Example: 2017-05-02T15:35:01.560Z
    • Addition CLOCK_SYNC as a reader health code, in the Reader Health category ClockSyncStatus
    • Addition READER_OUTDATED as a reader health code, in the Reader Health category SoftwareStatus
    • Addition of the following fields to the Software Upgrade response: created, updated, numDevices, isCancelled, numFailures
    • Addition of the SKIPPED value to the Software Upgrade show status endpoint
    • Removal of connectionType from the Reader Definition API API - this was deprecated in 2016r6 but is now completely removed
    • Require agentIdentifier in Reader Definition API - in 2016r6 this attribute could be null but now must have a valid string value
    • Removal of the value NORMAL from the valid list of values for the operation field in Reader Configurations - this value was allowed but unused in 2016r6 but is now removed
    • Removal of the field reportConfig in Reader Configurations
    • Additional check added to ensure that only one of channelConfig or txFrequenciesInMhz is specified in Reader Configurations


  • Simplification of operation-specific reader configurations and recipes, for location and inventory
  • Item absence detection calculated via tag expiration
  • Individual Reader Health, available via API query, AMQP messages, or SNMP traps
  • Bulk and individual reader Firmware Upgrades
  • Reader agent installation via the ItemSense Management Console, to enable improved reader-based location calculations, reader health monitoring and firmware upgrades
    • Readers no longer interface to ItemSense via the LLRP channel, and the reader agent must be installed to interoperate with ItemSense
  • Delivery and installation/upgrade of ItemSense binaries via installation script and docker containers
  • Support for offline installation and upgrades
  • New APIs
  • API changes
    • Additional Reader Configurationsoperation values: INVENTORY and LOCATION, with operation-specific configuration
    • Additional Recipestype values: INVENTORY and LOCATION, with type-specific configuration
    • Additional reportToHistoryEnabled parameter to the Jobs API
      to allow performance improvements when reporting via Item History is not required
    • Removal of connectionType from Reader Definition API
      Note that if this parameter is set, it is ignored.
    • Removal of unused labels field from Reader Configuration API
    • Addition of internal-use-only agentIdentifier parameter to Reader Definition API
      Note that this parameter is used internally by the ItemSense Management Console to provision reader agent software.

In addition to the above features, there is an architectural change. Along with the existing ItemSense software components, the ItemSense Core and the ItemSense Management Console (IMC), there is a new component installed onto each reader being managed by ItemSense. This reader software component is called the Reader Agent. It is the reader agent that makes Health monitoring and Software management possible. The reader agent installation managed by the ItemSense Management Console.

The addition of an Impinj-only agent on the reader allows for a custom communication protocol which offers a number of advantages. Namely: * Enhanced Security – Authentication and encryption of the communication between the reader or gateway and the ItemSense Core. * Easy setup - As LLRP uses a port that is typically blocked by firewalls (port 5084), the customer protocol uses HTTPS (port 443) which is typically an allowed port but corporate firewalls. * Additional control points - The Reader Agent allows additional logic to be placed on the reader. For example, unique reader-based custom algorithms for gathering Item Intelligence from Tag read data. * Additional reporting data  - This could be data about the reader itself. Reader health information is a great example of this. * Additional reporting modes - This allows for various ways to aggregate the data on the reader itself before the data is sent to the ItemSense Core component.

August 4, 2016

  • Improvements to IMC:
    • Jobs can now be set to run indefinitely
    • Reader Definitions now fully support xPortal Gateways
    • Reader Modes are now filtered according to the regulatory region set during ItemSense installation
    • Improvements made to Zone Map editing including the ability to edit previously created Zones.

July 5, 2016

May 2, 2016

  • Get Item History now includes a new parameter alwaysIncludePageMarker that will always return a page marker so you can retrieve all new records since the last time you queried
  • Message Queues can now be configured for item location movement with filters for minimum distance moved
  • Configuring a Message Queue now has a parameter called zoneTransitionsOnly that filters for item movement between zones.
  • Get Item History and Message Queue now return floor changes.
  • Tag identifier (TID) is now populated from Get Items, Get Item History, and Message Queues
  • Zone Maps now support polygonal and non-contiguous zones
  • Support for reader modes 1003 and 1004 in Reader Definitions added
  • xArrays can now be used in Recipes set to GATEWAY zoneModel
  • Improvements to IMC:
    • Support creating polygonal zones
    • Improvements to editing Antenna Zones for speedway readers
    • Added unit of measure
    • Improvements to the display of job errors
    • Start delay on Jobs is now supported
    • IMC can now be updated along with ItemSense

March 8, 2016

  • Added filter to Get Items to filter by last modified time of the item
  • Added support for GATEWAY zoneModel for xArrays tagId is now populated again in item queries (when tidIncluded is enabled)
  • Get Item History now reports all item movement history in addition to item movement between zones
  • Zone Transition Messages now include item location coordinates (x, y)
  • Jobs will now start when some readers are disconnected
  • Placement config data no longer required when adding a reader definition when running readers gateway mode
  • Floors are now String data type rather than Number
  • ItemSense service updates now deploy with minimal interruptions to running jobs
  • Small bug and usability fixes to the IMC
    • Fixed an issue where zone points would not accept 0 values
    • Fixed an issue with sessions expiring too quickly

January 27, 2016

January 6, 2016

  • Improvements to IMC:
    • Ability to set and clear the current Zone Map
    • Added ability to install IMC on existing ItemSense Core installations
    • Enhancements to stability and user friendliness of ItemSense Core installer
    • Migrated to Node version 5.3.0
    • Fixed an issue where changes to Reader Configurations would not immediately be visible in the UI
    • General bug fixes and stability improvements

December 17, 2015

  • Initial release of the IMC (ItemSense Management Console), a web interface for ItemSense, with the following initial features:

December 10, 2015

  • Floor names can now be strings
  • ItemSense service upgrades without without interrupting job execution
  • Improved Get Items performance
  • Improvements to location computation and floor assignment