iMouse: Modular and scalable 24/7 live monitoring

The challenge

In laboratory animal husbandry, it is common practice to continuously monitor and evaluate the behavior, responses, and weight of laboratory animals. In most cases, these human interventions do not correspond to the circadian rhythms of the animals. The daily in-eye monitoring of laboratory animals and the documentation of the resulting data is regulated and required by law. In order to promptly notice a possible deterioration in animal welfare and to be able to intervene if necessary, animals undergoing stressful experiments must even be subjected to in-eye examination several times a day. Each of these interventions triggers a stress situation throughout the home cage. This has been shown to affect animal behavior, general well-being, and metabolism to the point of phenotypic outcome, even when the data collection method is automated.

In addition, the common method of monitoring provides a snapshot of the animals housed in the home cage. The overarching goal of the project is to provide stress-free and continuous monitoring of the experimental animals in their respective home cages, which is also consistent with their circadian rhythms.


The solution:

The IIoT-Guidance and IIoT-Projects team conducted several workshops and interviews regarding the use of the cage systems and necessary optimizations.

The desire to be able to observe the test animals remotely, with multiple participants, from multiple perspectives was included. Also, the technical solution should be cost-effective and understandable.

One of the biggest challenges is a secure integration into the existing IT infrastructure (if available), which is why a WiFi data transmission was chosen at the beginning. The next step was to switch to a LAN connection in order to be able to transport the necessary data.

In order to keep the system simple, the data is transmitted via cable.

Video signals are transmitted together with sensor data to a simple compute unit and from there to a local server owned by the customer. 

Access to the local server is possible via a secure and target group-specific VPN connection with a personal key. The infrastructure belongs to the customer.

It is possible to export and process the data via the cloud.

AI models can be trained with the exported data and updates to the algorithms can be fed back to the system. There is still great potential for optimization here in the future. Running AI models directly
on the system was not implemented in the first step for cost and performance reasons.

On-site assembly was handled by IIoT Projects. Customer service and support will also be provided by IIoT Projects, mainly online.

The Europe-wide roll-out is planned via partners. The optimization of the system and the adaptationto other cage systems is constantly taking place.  to other cage systems is constantly taking place. The algorithms to be developed are also in focus.
Here, close cooperation with research institutions and AI experts is taking place. Further partnerships are being formed in order to be able to monitor further animal species and to constantly
further development of the system.

On-site assembly, including IT integration, is provided by IIoT-Projects. Customer service and on-site support are also provided.

The further development of the basic functionality is done in close cooperation with the customer. Further functionalities are developed at the request of the customer, algorithms are implemented to be able to generate automatic alarm messages in the future. Until we get there, a close exchange between data analysts, animal caretakers and programmers is "daily business".

The result:

The IIoT-Guidance and IIoT-Projects team, together with partner HPI, has developed an end-to-end automated monitoring system for laboratory animals housed in standard cages. This system integrates completely with or does not change existing workflows at VTH, and the low hardware costs can be passed on to the customer, funding the R&D team to expand the portfolio into other project components such as automated mouse scales and connectivity to the local ERP system. Furthermore, the iMouse DIGI Frame can be used as a standardized data mining platform across the board and for training an AI where further research is done with partners. For this purpose, the iMouse TV (CCTV and data agent platform) serves as the technical foundation, a scalable open hardware and software solution. By using easy to integrate camera modules, especially the service becomes easy and cost effective to realize compared to other solutions.