04 Sep Automatisierung Tierraum
Basis for discussion
A shortage of skilled staff and increased demands on research and animal welfare require a rethink of processes in animal rooms. The shorter the time spent by people in an animal room during the sleeping period for rodents as nocturnal animals, the better.
In modern laboratory animal facilities, the automation of the animal room is therefore playing an increasingly important role. Looking at the animal room from this perspective, the two “machines” that can be automated in the animal room in terms of handling, communication, documentation and animal welfare are the rack and the changing station. These two objects are closely linked in the day-to-day operations in the animal room through the cage changes and are also moving ever closer together from a technological perspective.
Extensive processes of license management, planning for experiments, or the planning of breeding and procurement take place in advance outside the animal room in the animal management system. This data now manifests itself for the first time in the animal room on the cage card when the cages are initially stocked with animals. From now on, the processes of changing the cage contents begin, which primarily take place at the changing station or medical workbench. The effects of these processes must be monitored regularly and closely as soon as the cages are in the rack.
Automation and the seamless interaction of the 3 components animal management system, changing station and rack when observing the animals in the rack and when manipulating the cages and animals on the workbench ensure improved data quality, lean processes and higher animal welfare
In detail, these are the following technical improvements:
- Daily cage check versus immediate information by signaling anomalies
- General regular cage changes versus targeted cage changes
- Extensive personal observation versus continuous automated observation
- Evaluation of personal recording versus evaluation of automatically recorded data
- Searching for cages versus signaling cages with the search criteria
- Manual or semi-automated focusing (QR code, Rfid) on cages on the workbench in the software versus immediate and simultaneous presence of all cages on the workbench in the software after pulling the cages in the rack.
- Extensive navigation in table-oriented software for the documentation of processes versus simple confirmation of preemptively suggested processes based on the cage and animal data called up
- Input errors versus avoidance of errors due to rule-based semantic checking
- Manual creation and change of cage cards at the cages versus automated creation of cage cards directly at the cage
- Searching for the position of the cage in the rack versus the position displayed on the cage card
- Mixing up of cage cards and cages during exchange versus permanent presence of the cage card on the cage
- Unnecessary routes with manually created or printed cage cards versus only necessary routes with cages between rack and changing station
- Frequent disinfection of contamination bridges such as touchscreen, mouse or stylus versus reduction of contamination bridges and simple disinfection from the workflow
- Complex and detailed planning of cage transactions versus simple free text planning and detailed execution with the help of the expert system
- Long periods in the animal room during the rodents’ sleeping times versus short periods due to planning and digitized preparatory work in the office.
Simple spaghetti diagrams can be used to determine the savings potential from pulling the cage in the rack to opening the cage at the changing station.
The locations of the racks, changing station and all computer workstations for recording and maintaining animal data are drawn to scale in a room plan.
Red lines represent routes with paper cage cards, black lines represent routes with cages. The path lines are provided with lengths and times.
CageTalkers automatically update themselves directly at the cage, thus eliminating all paths with paper cage cards, which are shown as red lines.
By using the DVC in combination with the VCS, all times of the paths marked in black are reduced.
In an extended representation, this can be shown for each process from pulling the cage in the rack, including animal manipulation with the creation of the cage card, to putting the cage back in the rack, thus demonstrating the benefits in detail.
Spaghetti diagram
Interfaces:
Of course, such increases in workflow efficiency are not only achieved through the automation of the individual “machine”, but above all through the exchange of data between the systems via interfaces. These interfaces can be implemented today using SOAP and REST protocols.
Below is a description of the interfaces already implemented today. The illustration [Software Transactions] shows the user interface of the system used at the changing station. Depending on the system used, this may be VCS, AMS or DVC Operator.
It is assumed that the operational data of the cage changing, and animal manipulation processes are recorded and documented where they are generated, i.e. at the changing station, and are not noted down manually and then added later.
This operational real-time data must be distinguished from the more administrative and planning data, which is also reflected on the cage card but is not generated in the animal room. In a hospital, this would be referred to as medical and administrative records.
The interfaces available today enable a complete closed technical cycle of the cage and animal data within the CageTalkers, DVC and AMS systems.
If the AMS is used directly at the changing station instead of the VCS, the DVC interfaces 3, 4 and 13 are displayed in the AMS instead of the VCS Control.
Scenario 1 CT_DVC interfaces (CageTalkers are available DVC will be retrofitted)
1. AMS creates cage card via Cagetalkers web service
2. Cagetalkers web service updates Cagetalker
3. Cage for transferring the cage card data to is selected in the DVC browser (operator)
4. The cage data of the initial placement is transferred to the DVC via the AMS CageID and TierIds
5. One or more cages are pulled from the DVC rack and communicated to the DVC master
6. The DVC master transfers the drawn cage numbers (and other measured values) to the DVC server
7. The DVC server informs the Cagetalker web service of the drawn cage (and other measured values) as floating.
8. The floating cages are either opened automatically (a CS in the room) or opened with the help of the DVC reader or VCS reader
9. The cage with all its data is displayed in the Virtual Changing Station (VCS). Various functions for manipulating the cages are offered
10. Changes to the cage contents are made on the changing station and documented in the VCS
11. The VCS updates the Cagetalker web service
12. The Cagetalkers are updated
13. The DVC server is updated
14. The AMS retrieves the changes based on CageID and TierIds
Steps 6 to 9 are carried out seamlessly in the background, i.e. not visible to the user.
Scenario 1 Simplified without VCS Control and DVC Master:Scenario 1 Vereinfacht ohne VCS Control und DVC Master dargestellt:
1. AMS creates cage map via Cagetalkers web service
2. Cagetalkers web service updates Cagetalker
3. Cage for transferring the cage card data to is selected in the DVC browser (operator)
4. The data (initial placement) is transferred to the DVC via the AMS CageID and TierIds
5. One or more cages (with measured value data) are drawn from the DVC rack
6. The DVC (master and server) informs the Cagetalker web service of the cages drawn as floating cages.
7. The floating cages are either opened automatically (a CS in the room) or opened with the help of the DVC reader or VCS reader
8. The VCS updates the Cagetalker web service
9. The Cagetalkers are updated
10. The DVC is updated
11. The AMS retrieves the changes based on CageID and TierIds
This interface architecture manages with only one outbound connection via the Cagetalkers from the AMS and one return connection from the DVC to the AMS for 2 systems. This means that the black dotted connection, i.e. the outward connection from the AMS to the DVC, can be omitted when using Cagetalkers. If changes are made to the cage contents with VCS, the VCS updates the Cagetalkers directly and the DVC’s back interface to the AMS is used to transfer the data changed by the VCS to the AMS. In terms of data technology, this is therefore a fully automated cybernetic closed loop.
Scenario 2 DVC_CT interfaces (DVC are available, CageTalkers are retrofitted)
In this case, the CageTalkers would merely be an appendage to the DVC. However, three interfaces would then be used. On the one hand, the back-and-forth interface of the AMS to the DVC and additionally the interface from the DVC to the Cagetalker web service. From an interface perspective, the DVC assumes the role of the AMS regarding the Cagetalker. This approach does not quite correspond to reality, because the initial stocking of a cage in the DVC always takes place with animals on the cage card generated by the AMS. This link via the cage number and the animal numbers of the AMS is necessary so that the DVC can receive and process data from the AMS at all. A paper card with AMS cage ID would therefore first have to be created and the animal data would then have to be transferred to the DVC on this basis. Only then could the DVC continue working with Cagetalkers.
1. cage for the transfer of the cage card data to is selected in the DVC browser (operator)
2. the data for the initial loading of the cage is transferred to the DVC server via the AMS CageID and TierIds
3. the DVC server transfers data to the Cagetalker web service
4. the Cagetalker web service creates a cage map
5. DVC Rack informs the DVC Master of the cage status (pulled/in slot and other measured values)
6. the DVC master informs the DVC server of the cage status (pulled/in slot and other measured values)
7. / 8. back and forth connection AMS and DC for data exchange via the AMS CageID and TierIds.
There are some important considerations to improve the hygiene of the input medium for the AMS software, the DVC operator or the virtual changing station VCS and the automatic detection of cages on the changing station. Contamination bridges should be limited to the unavoidable contact with the cage lid, the cage tray, the water bottle and the animals. Ideally, there should be no contamination of the input medium at all. This could only be achieved with voice recognition, gestures or eye control. However, our current development focus is on input within the LFS barrier, minimizing the number of objects required for input and the surface area to disinfect the input medium as quickly as possible during the essential disinfection of hands after each cage change.
The following options are currently state of the art.
The AMS software and the VCS control software are generally operated via touchscreen, mouse or stylus. All three elements represent contamination bridges and must be disinfected after each cage change. Cage detection is semi-automatic with a scanner, which must also be disinfected each time. This contamination bridge can be eliminated if the Galilei Talkits and a VCS RFID reader strip or the DVC cage trays with RFID readers are used as an alternative and cage detection is therefore automatic.
In addition to the software under Linux, the DVC operator consists of the DVC trays for the cage lids, each with integrated RFID readers and a computer with a 15 x 15 cm touchscreen on a footprint of Xxxc. This device is easy to disinfect. Readers with input currently occupy a footprint of Xxxc.
The Galilei Stream Deck Flow is a slanted, easily disinfect able keyboard for the laptop with 3 x 4 backlit haptic keyboard keys on a footprint of 10 x 12 cm. The keys update dynamically with the same icons as the VCS application buttons and provide support for all animal transactions with minimal navigation in the software due to the integrated expert system. With only one changing station in the room, the cages automatically appear in the VCS software with pulling the cages from the DVC rack. If there are several changing stations in the room, the DVC readers or the VCS reader bar are used to recognize the cages or to manually select the floating cages.
Alternatively, racks could also be linked to a changing station.
Below are some videos to illustrate selected processes in scenario 1
The cage card for the initial loading of the cages is created using Microsoft Excel instead of an AMS.
- Initial equipping of the cages with Excel and mating with creation of Cagetalkers and linking of the DVC slots
https://app.screencast.com/y0XtE1Vm2rhjg
- Mating via Streamdeck Flow
https://app.screencast.com/unO2txxq0Lxck
- Discharge cages and weaning via Streamdeck Flow
https://app.screencast.com/L7xRDyZ928XOX
- More examples to follow.
Conclusion
The interaction of Animal Management System (AMS) with CageTalkers, Virtual Changing Station (VCS) and Digitally Ventilated Rack (DVC) closes the cybernetic circle of automation for cages and animals in the animal room.
As automatable “machines”, the changing station and rack play a special role in streamlining processes. Not only is medical data such as activity, aggression and other behavior measured and signaled in the DVC rack, but logistical processing is also improved and presented to the outside world in a sustainable and evaluable manner with the help of integrated interfaces and automation using the semantics of cage data in the VCS. In this concept, the AMS primarily assumes the role of the administrative part of the application, the management of licenses, the procurement of animals, experimental planning, the initial stocking of the cages, billing and reporting. Data exchange is secured via the DVC/AMS interface.
In addition to saving time, the focus is on improving the quality of data for research and reducing and refining animal welfare.