Autonomous Track Missing
Introduction
The Autonomous Track Missing is a read mode specially designed for Lift & Learn and similar systems. Systems driven by user interaction, that require fast and reliable product removal detection.
The Autonomous Track Missing read mode works as follows:
Performs a calibration period that can span up to a few seconds.
Keeps a list of continuously read tags.
Generates an event when one of the controlled tags has stopped from being read (TAG MISSED).
Generates an event after a MISSED TAG is read back (TAG BACK).
Generates an event when there is a new tag that was not in the controlled tag list (TAG NEW).
The Autonomous Track Missing is generally used in Lift & Learn systems, where the user interaction requires the following characteristics:
The RF detection is uniform across the work surface
The RF detection vanishes very fast, usually less than 40 cm.
The above guarantees a very powerful effect, but is a challenge for RFID hardware.
False positive product removal
One of the common problems in Lift &Learn systems are the false positives: false product removal events.
How good a system performs is measured in terms of false positives per day.
Another important characteristic of these systems is the reaction time: how long it takes to detect product removal.
In Lift & Learn systems we must reach a compromise between reaction time and false positives per day. The shorter the reaction time, the larger the false positives per day, and the other way round.
It is generally accepted a system with a rate of false positives per day beyond 2.
Possible ways to improve the false positives per day rate are:
Increase reaction time. Note this has a direct effect on user interaction time.
Increase read power. Note this may increase removal distance.
Increase sensitivity. Note this may increase removal distance.
Multi-tagging: tag items with more than one identical tag.
Recommended hardware
Any RFID hardware can potentially be used in a Lift & Learn or similar system. However, to achieve the best user interaction it is recommended to use RFID near-field hardware:
Near-field antennas: antennas that have a magnetic radiation component in the near-field region.
Near-field tags: loop tags that respond basically in the near-field region.
When used in the right way, near-field hardware will provide the two user interaction required characteristics.
Regarding tags, it is known the tags below have proven to work in Lift & Learn systems:
Smartrac Trap NF
Trace Tech TJ24 Blade
Unfortunately, the final configuration of antennas, tags, products, tag placement, etc must be tested thoroughly. Any change on a Lift & Learn system may result in an incorrect behavior.
Note of caution
The track missing mode will only be successful when the detection time is around 1 s or less. Achieving such level of reaction time is very challenging. To achieve such level of reads the Track Missing mode requires certain conditions:
Limited number of antennas: we suggest starting with 4 or less antennas
Limited number of products: we suggest starting with one product per antenna
It is recommended to use confined antennas like Advantena-cp-11
Sources of potential problems:
Change antenna types
Change antenna locations
Add or remove antennas
Change the tag type
Change the product type
Changing the number of tracked products
Etc
Any time one of the above items changes, the system may require a nee evaluation of:
Tag type
Tag placement
Antenna placement
RF power and sensitivity
Etc
Configuration
First select Autonomous mode and then select Autonomuos Track Missing
How it works
The operation is as follows:
Starts reading for a Calibration time (ms) to create a list of controlled tags or calibrated tags.
After creating this list, it will check for each tag:
if a tag has not been read for more than Missing time (ms), it will create a tag missed alarm.
If a tag was considered missed and has been read again, it will create a tag back alarm.
If a tag has been read and it is not in the calibrated tags, it will create a tag new alarm.
After the calibration refresh time (s), it will start again the operation.
If the Dynamic Calibration is enabled, there will be no calibration time or calibration refresh, every tag will be added to the list of controlled tags or calibrated tags as is being read.
XML Messages
TAG_MISSED
ADVANNET/1.0 Content-Length:520 Content-Type:text/xml <?xml version="1.0" encoding="UTF-8"?>
<deviceEventMessage>
<type>deviceEventMessage</type>
<ts>1473752611907</ts>
<status>OK</status>
<event>
<class>com.keonn.device.impl.DefaultTagAlarmEvent</class>
<type>TAG_ALARM</type>
<typeclass>com.keonn.spec.event.DeviceEvent$EventType</typeclass>
<deviceId>AdvanReader-m4-150</deviceId>
<alarmType>TAG_MISSED</alarmType>
<epcs>e2006b060000000000000008</epcs>
<readerPort>0</readerPort>
</event>
<deviceId>AdvanReader-m4-150</deviceId>
</deviceEventMessage>
TAG_BACK
ADVANNET/1.0 Content-Length:518 Content-Type:text/xml <?xml version="1.0" encoding="UTF-8"?>
<deviceEventMessage>
<type>deviceEventMessage</type>
<ts>1473752613704</ts>
<status>OK</status>
<event>
<class>com.keonn.device.impl.DefaultTagAlarmEvent</class>
<type>TAG_ALARM</type>
<typeclass>com.keonn.spec.event.DeviceEvent$EventType</typeclass>
<deviceId>AdvanReader-m4-150</deviceId>
<alarmType>TAG_BACK</alarmType>
<epcs>e2006b060000000000000008</epcs>
<readerPort>0</readerPort>
</event>
<deviceId>AdvanReader-m4-150</deviceId>
</deviceEventMessage>
TAG_NEW
ADVANNET/1.0 Content-Length:519 Content-Type:text/xml <?xml version="1.0" encoding="UTF-8"?>
<deviceEventMessage>
<type>deviceEventMessage</type>
<ts>1473752600660</ts>
<status>OK</status>
<event>
<class>com.keonn.device.impl.DefaultTagAlarmEvent</class>
<type>TAG_ALARM</type>
<typeclass>com.keonn.spec.event.DeviceEvent$EventType</typeclass>
<deviceId>AdvanReader-m4-150</deviceId>
<alarmType>TAG_ADDED</alarmType>
<epcs>e2006b060000000000000008</epcs>
<readerPort>0</readerPort>
</event>
<deviceId>AdvanReader-m4-150</deviceId>
</deviceEventMessage>