The year 2021 was thrilling – even without the global pandemic situation.
The year started, like 2020 ends….with Gaia-X.
The year in details
In 2020, I was responsible for managing the WS 2(technical implementation) within the GAIA-X project. From 1/2021 my focus was to develop of the community platform concept for the GAIA-X AISBL, as a contractor for the BMWi.
Besides this innovative European project, I became a part of another disruptive and existing idea within a newly established team. This idea is based on the backbone of research, of medical development, focused on small animal. The mouse. The hero in translational research.
The two ideas or better projects doesn’t interconnect, since the mouse as experimental model has not the political standing as to ensure the safety of private data. But, if we take a more precise look, than both ideas are focusing on data.
Considering, that scientific results need to be validated in a living organism and moreover the development of medical applications need to be tested at least in mice (by law), we came to the question, if we, as humans, could reduce the level of intervention and subsequently the level of stress for our small friends.
So, the idea was to remotely monitor laboratory animals without human interaction.
Since the idea is, to record these animals while they were employed in experiments for evaluating medications, used in human as well as in animal health care. Moreover, the vision is to use or even share these data within the research community.
The questions was: What systems are available at the market? Are they affordable and moreover scalable? What about the data ownership of the acquired data? Is the institute the owner or the data?
Having read several research reviews and company pages regarding the monitoring and tracking of laboratory animals, we decided to create a generally simple system which could be used as retrofit solution. Means, we wanted the system to be integrated in the existing facility environment, regardless of the cage provider. Means also, the system should be scalable and very important also affordable.
We named it iMouse, stands for Integrated monitoring & operation system in use within a standardized environment.
Since the Fundament of iMouse is following the open-source and open-science spirit, we build up a strong and reliable partnership with the Heinich-Pette-Institute (HPI). So, the projects became more physical and in 3/2021 we implemented the first PoC in the animal facility of the HPI, headed by Dr. Oliver Strauch. Here we combined the selected open source software with simple hardware available nearly everywhere for an affordable price. After the integration of the now named DIGI FRAME (because we are not building cages) in one existing animal subunit, strongly supported by Oliver Strauch, his team, and the local IT unit, we generated an interconnected PoC with remote access.
iMouse DIGI FRAME with DDC (digital data card) accessory
Having that first PoC up and running, we decided to follow the “Call for participation for lighthouse projects for GAIA-X, by the BMWi” and brought together a high-potential and GAIA-X experienced consortium to participate. We successfully developed the idea with a proven project plan to the next level. Unfortunately, our idea wasn’t selected as part of the lighthouse projects. It seems that improving the work with the small hero in life science isn’t prestigious enough, even if we have a strong open-source, open-science and AI development focus.
Anyway, we believed in the project and since its highly disruptive and has the potential to boost the pre-clinical research and development market, we decided to follow the “SPRIN_D” call (https://www.sprind.org/de/) in June.
But again, after the first project document submission and interview, we realized the “SPRIN_D” governmental agency wouldn’t found the project. One crucial factor was the missing but required IT-infrastructure.
At this point, we realized, to fully establish the iMouse project (incl. data exchange platform), we need a GAIA-X equal solution to exchange (raw) data in a safe and trusted way for the different target groups. In consequence, we applied to be part of the health domain at the GAIA-X German hub (link: de, eng).
Again, we followed the potential of the project idea and installed further PoC`s with a different camera setup and in a higher security environment (S3) with hamsters to show the potential is not just stopping with mice.
In June, the IIoT-Projects GmbH was born. The IIoT-Guidance GmbH developed the idea and brought the right partners together but it was time bring it on a more experienced and stronger knowledge base. The IIoT-Projects GmbH therefore works as an incubator, to develop PoC`s and subsequently a real solution for the industrial environment. The first action was to realize the implementation in the research field for the iMouse project.
founding team of IIoT-Projects GmbH
(left to right – Mathias Haensel, Dr. Janine Kah, Mirko Lampe, Stefan Koeppel, Nader Halmuschi)
To succeed with our project, we need to evaluate moving pictures, developing and train algorithms to upgrade simple data to added value information what finally enable the generation of statistical data and alarm signals to deliver an additional value within the research landscape.
Therefore, an interconnection of different specialists set the fundament for the iMouse project. We established partnerships in the research landscape, had intensive discussions with market companions, R&D-, animal behavior- and AI- specialists. Partners are testing our developed solution to improve systems stability, usability and finally functionality.
This interaction, as mentioned before, is essential to develop the proven working idea to a working, added value information system!
Nevertheless, the iMouse team developed out of Janine Kah’s idea – a solution which solves already today a market need. The need at the beginning was, having the possibility to see how the wellbeing of the supervised animals is, within an existing home cage they are living in.
For the GV-Solas conference on 24.09.2021 (59th Annual Meeting of the Society for Laboratory Animal Science) we applied with the iMouse use case and we got selected as an important digitalization milestone. Janine presented the development progress of the project, and we first time presented our PoC of an automatic scale system (as part of the iMouse approach). The uniqueness here is to acquire mice weight within the home-cage without human interaction.
The iMouse scale contains of a tube unit which works also as enrichment (compare figure below). Every run-through, result in a new data point which can be recorded and extracted for a gapless measurement within a certain experiment.
These data points were recorded in real time combined with cage specific temperature, humidity and exact time stamp. All these information get integrated in the recorded material. Today we distinguish the 4 to 5 individuals in the home cage by their ear – and tail marks, recognized by the review of the recorded material. The next step is to extend this approach. Firstly, by sending this particular data to the ERP system, for each individual by using a bi-directional API. Secondly by collaborating with an AI specialist to automate recognition within the recorded material.
iMouse scale prototype in practical use within the iMouse DIGI-FRAME
On the 1st of December 2021, after a year of development, we finally sold our first iMouse system.
Our open source-based iMouse system is developed together with the customer, retro-fitted around existing cage systems, scalable, and most important, the customer owns his data.
With our rental sales model, we offer a full-service package. This gives our clients transparency about the costs. The customer just pay a one-time installation rate to interconnect the iMouse system to the local IT infrastructure, depending on the existing IT infrastructure and monthly expenses for the access of the DIGI-FRAME and data recorded. We choose this model since further software functionality and upgrades will improve the existing system over time.
Many companies are working on a digital cage solution. Most of these missed the required of scalability, the solution was not retrofit based and it is over affordable budget.
A lesson learned for our team: There are no analysts available at the market, and before the system is working automatically it must be trained by skilled analysts working together with animal behavior specialists.
That’s the reason why we call for partners. We are looking for development partners, annotation specialists, data scientists with video-/ picture background and developers to make our vision happen.
To sum up, we like to further expand our network for the iMouse projects. Offering special benefits like a system installation without or at least lower costs would be a part of the partnership. The benefit level will depend on the support a partner could provide to improve the project. And finally the biggest advantage is to be a part of the future.
Here we are, looking forward for your feedback and many thanks to all the discussions we had in 2021 and to our team: Janine Kah, Mathias Hänsel, Nader Halmuschi, Stefan Köppel, Alexander Banas, Hendrik Krüger
Many thanks to all our supporters and discussion partners!