Can you film at night/in the dark cycle?

Our first system does not allow video monitoring at night. We are currently working to integrate night vision into our system. We use typical night vision wavelengths that shouldn't be visible by mice. We can control when the lights come on so that they're not persistent. And we have cameras that can visualize the environments with infrared light all night.

How are the animals' activities tracked and what information does the data provide?

Activity data is collected by computer vision techniques that we're using to look for movement, changes in the field of view/frame. For now we are looking simply at movement. In the future, we'd like to track different types of movement, number of mice, types of activities, and so on. Right now our graphs show you cage level movement as it changes across time. You can compare movement today to movement in the past and see the overall level of movement in the cage and how it may be changing. It is cage level, vs individual mouse movement, for the purposes of daily cage inspections. For scientific purposes we're working on this individual level animal movement in our SBIR projects.

Can you use any type of rodent rack system?

Our equipment works on most normal shelving unit style racking systems such as Allentown, Tecniplast, Innovive, and Animal Care Systems.

Do we need to have a wired network connection?

We can do both wired and wireless connections. To save costs, we make multiple cameras connected to a single computer. That computer is a mini pc, 5" x 5", and it can communicate wirelessly or by a wired connection to the network. We can customize this based on your needs, space, budget, etc.

How much data does say 25,000 cages create? Can existing networks handle that level of data if all cages had cameras?

We use local storage on powerful mini pcs to do most of the processing and filter the data. We filter out the irrelevant data, and only send a small subset of data from the PCs to the network or cloud. Therefore, we're transmitting and storing much less data than you would expect, and your existing networks should be able to handle it.

Are facilities replacing in-person checks with solely remote monitoring, such as on the weekends? Do the regulatory bodies accept this over in-person checks?

We reached out to the National Institute of Health (NIH)'s Office of Laboratory Animal Welfare (OLAW) regarding this topic. They said that Institutional Animal Care and Use Committee (IACUC) decides if this is acceptable or not. From their opinion, they say that it is acceptable, especially where there are other issues such as in ABSL3+ environments, to allow more frequent checks of animals, while minimizing exposure to humans. And, we believe that IACUC would agree that weekends are a good use of this technology as well, especially for improved monitoring of higher risk animals.

Can inspections be done automatically, i.e. where animals are flagged if there is a problem?

This feature is part of our upcoming SBIR grant funding work. By September 2021, we expect to have preliminary data on automated health event notifications, where there is an abnormality in the way that the animals are behaving in a particular cage. We have lots of methods that we plan to employ to accomplish this feature during the project period. If you are interested in trying the system at your facility as part of our study, please contact us to learn more. We want to include different settings, facilities, cages, bedding types and all of the other variables during this study.

Do you have automated detection of pups?

No, this feature is not yet automated, but it is on our roadmap. The more interest there is, the sooner that we will develop it. Some of our existing algorithms may be able to help with this automation by Q1 2022.

Are there automated alerts or is that coming, e.g. for the cage that has not had any movement today, etc.?

If you are interested in automated alerts and have good study sets that you believe would yield good data for this effort, please let us know. We'd be happy to help install the system and let you use the system as part of our study to gather data. You could then keep the system after the study with ongoing subscription fees. http://swiftagsystems.hubspotpagebuilder.com/learn-more-about-swiftag

Will this system be covered via an IACUC Policy that discusses telehealth?

Yes, definitely. IACUC is starting to have these policies and we can continue to work on it, so that veterinarians on the weekend can begin to assess a situation remotely from home to be proactive. If husbandry staff alerts you to a health event, the veterinarian can begin the assessment from home and determine if it can be addressed remotely, or if you need to be there in person to address it. It will save some trips to the facility for non-urgent issues, especially on weekends.

What is the precision of your event detection for different cage scenarios?

This is part of our study - types of bedding, nesting materials, strains of mice, etc. we want to be able to be versatile to cover all of these variables. We want a more generalized algorithm to assess overall health.

I saw on one of your slides, 8x USB, to computer, does that mean that for every 8 cages there needs to be a separate computer, or is there some way computers can switch to different sets of cages, etc.?

Yes, 8x USB to a computer. We string the cameras together to reduce cost and data usage. If you need activity data for study purposes we can provide a computer for each camera, but for daily inspections that isn't necessary.

How does the clarity of the cages affect video quality?

It can be dramatic depending on how old the cage is - uv damage and washing can make the cages difficult to see through. This is the same as humans looking through the cage vs a camera - same limitations. We recommend using new, clean cages when using our system. We have considered creating disposable cages as part of our system, but that isn't something that we currently provide.

With the system as it is now, how would you know that the motion is decreased from yesterday. There were charts below the camera images, are they automatic at this point, and do they show changes in movement over time?

We do this now by monitoring data continuously throughout the day - any activity or motion events through the day and normalizes it for the number of animals in the cage. What you see is aggregate motion data, and then you can compare it to the rest of the rack. So you can see how each cage reacts to changes in the environment such as a pressure change to a door opening or someone entering the room. If animals don't react, that is an important observation. We can add things on here in the future.

How easy is to attach your camera to the cages? How long does it take to take the cages out for cage washing? Does that itself add up to the overall operation?

It is very quick and easy to re-attach cameras after cleaning. It doesn't add up to any significant amount of time. Once every 3-6 months when the racks need to be washed, you can remove the cameras and put them in the fresh rack, and that takes just as much time as removing the cameras, cleaning and replacing them.

What is the max number of animals per cage and how do you distinguish between different mice that are the same strain?

We're not currently distinguishing between different mice by ID at the individual level. We know the number of mice and if they're from the same or different strains. We can also integrate our animal ID tags for this purpose along with the cameras to get individual level data. We're currently looking at cage level activity and data, vs individual animals, because you would expect cage level activity to change if there is one animal sick or three. If there are more animals, such as five, it makes it more difficult to determine at the individual animal level. Our SBIR will give us that number of animals that we recommend per cage.

If it is for an ABSL3+ area, do you need to access the area or we can set-it up ourselves?

Yes, we've done both. It is easiest for multiple racking systems that you've cycled. We install the system outside the room and then you place it in the facility after doing cleaning/decontamination. Newer system is easier for DIY install. We can provide remote support to help you do it if you want to install it.

What is the process for evaluation of equipment?

Complete the survey form on our website to learn more and we will contact you about trying the equipment as part of our research for future products, and we offer low cost trials to rent the hardware to test for a period of time. Discounted purchase options after the trial. We recommend 1-3 months, but 3 months is best. http://swiftagsystems.hubspotpagebuilder.com/learn-more-about-swiftag

How do you participate in new feature development?

Click Here to complete one or both of the brief surveys to sign up for trial periods of our current products or to partner with us for R&D of new products and features currently under development. https://swiftagsystems.hubspotpagebuilder.com/learn-more-about-swiftag

What is the pricing for SwiftSENSE Video-Based Remote Animal Monitoring?

Can you use your own Matlab, DeepLabCut, Tensorflow tracking algorithms with our system?

Yes, we enable custom algorithm upload and use on real-time data.

What framerate and resolution are the cameras?

Standard is 25 fps to 30 fps and 1080p, however, modules can be upgraded based on use case and request.