Needing help with unusual application

[WeatherMan]

Hello,
I'm hoping someone can help clarify a few hurdles I'm trying to clear before moving forward with a rather unusual project and application of an NVR and WIFI cameras.

I'm working on a project this summer where I'm building something called "Weather in a Tank", which is a slowly rotating tank of water that can recreate the fluid movement of Earth's atmosphere and various weather/ocean phenomena. The WiaT is used for outreach and education efforts. Here's a link to the project's initial designers over at MIT and a video of it working. It's outdated and I'm hoping my version is far more 21st century.

Because the camera is co-rotating, a slip-ring to protect the signal from EM interference must be used. Digital signal slip rings are expensive so the current generation of WiaT's all use analog video camera. It is my idea to use an NVR and WIFI cameras to solve this hurdle and make the leap into the latest in high quality video and recording.

Because I plan to use the project for mobile outreach (i.e. taking to schools, community events etc.), I can't always expect a WIFI network to be present. Before I move forward, I'm hoping someone can give guidance on the capability of if specific brand's NVRs to connect with WIFI cameras absent of a network. I plan to have a router for camera-NVR communication, but will this connection still work without internet behind the scenes? I know that removes the ability for remote viewing via browser or mobile app, but that's not necessary for what I need it for.

Thanks in advance.

 

[tangent]

What you need is a wifi router to setup your own network for the camera to connect to. But before you do that, the easiest option for a drop in upgrade would be one of the protocols that's evolved to send digital video from security cameras over coax like TVI.

Slip rings aren't that expensive if you go direct from china (just an example): Ethernet slip ring 1channel 100BaseT 6 circuits 2A+2 circuits 5A with capsule slip ring OD 22mm-in Cables from Consumer Electronics on Aliexpress.com | Alibaba Group
Cheaper ones may work. A camera shouldn't need more than 10mpbs, 2 twisted pairs.

 

[tangent]

Also basic ones on digikey (sparkfun and adafruit) and not too expensive: Structural, Motion Hardware | DigiKey
spec sheets imply they might work for ethernet. I'd twist pairs of wires you intend to use for ethernet.

if you do wifi, you might be able to use something like a raspberry pi running the network it could also display the video. A pi would also work with ethernet.

 

[WeatherMan]

Thank you for the response.We are designing and building from the ground up, so we are assessing which pathway is best and most cost effective (working with university grants). I had planned on having a stand-alone router for the wifi connection, but needed confirmation the NVR didn't require true internet. The TVI protocol is an interesting idea. We will need a basic slip ring for the power supply for the cameras anyway, so if we can get something that can pass through the cheaper slip rings like Coax, that will make things easier. It would also alleviate the concern about wifi latency on cheaper systems and ensures quality feed at all times. Would be great to get a POE system to reduce wires needed to pass through the slip ring.

The MIT build uses an AC motor with a potentiometer+cyclometer to control the RPM of the rotating assembly, which isn't precise and a little outdated. I have already asked my team to work in Raspberry Pi or an Arduino control for the motor and rotating assembly to precisely control the rotating rate to achieve the correct scales as well as give us a more up-to-date system and they have made good strides getting that side of things planned out. We briefly discussed the ability of the Pi to control the video feed side of things, but we weren't confident a pair of high resolution video feeds wouldn't tax the processors and introduce lag or latency.

Again, thanks for your input!

 

[tangent]

2 cameras is more difficult than one. You'll need to evaluate what bit rate and frame rate will be acceptable. Depending on what you do to the drive mechanism it could complicate using a slip ring for video as a motor could introduce more noise. 2 separate signals from 2 cameras will also increase odds of crosstalk issues if you try a slip ring. To me it seems like there are some parts of this that will require iteration to refine if more than one unit is eventually produced. If you can get enough funding and have enough time I'd experiment with several different options on the video front to decide what's going to work the best.

 

[fenderman]

Also note that using ip (particularly with wifi) will introduce some delay....1/2 second or more...if this is an issue stick with tvi/cvi....