Smart RV Capability

[the_hern]

Looks like there hasn’t been much movement I. This thread in a while. I’m picking up mu DSDP this summer and plan on going full automation where possible. I came across this thread in my research.

i am a fellow tinkerer, I’m in InfoSec by trade. I’d LOVE to find out more about how this has progressed especially any hardware tie ins as I was planning on a mix of Pi and Arduino to talk to systems. Please reach out to me!!

 

[flyboy013]

There hasn't been much movement on this thread, per se, but the capability continues to be improved. I finally resolved the issues that I had in interfacing with the KIB network and all is good!

I've have developed an Alexa Skill and can now do things like treat a shade as a shade instead of as a light bulb. ie. I can tell Alexa to open/close the shade as opposed to turning it on/off.

I'm in the process of designing a custom PCB which includes all the necessary components. I'm hoping to put a few beta units together and have then ready for the Newmar International Rally in July. I'm still trying to figure out how best to get this capability out there - work directly with owners or reach out to Newmar. There's pros and cons of each approach.

 

[redbaron]

If you need help with pcb stuff I have done a few.

Would be interested in your final product.

 

[BPeas]

I am also interested in your product, even assisting with testing. Would love to see it at the International.

 

[msramanujam]

I’d love to work with you to test things & also help make any improvements to the product. Please let me know how I can go about working with you. FYI, I’ve got a 2021 Newmar Dutch Star.

 

[smoorehome]

@flyboy013 I would be interested in testing your technology.

 

[DavidL]

Hi flyboy013 - I was wondering if you made headway. Didn't see you post on that "other" forum. So, tell us a bit more about your KIB adventure. Did you find protocol docs? Squeeze it out of KIB? Reverse engineer it? Are you talking to KIB via the RV-C network? Do they follow RV-C standards or did they mod it a bit to make it proprietary? Does your hardware talk only RV-C? or other networks as well? Does / Can your solution talk MQTT by chance to integrate with our myServer automation system? If so, we could very quickly add support for it. If not MQTT - any other integration methods supported? Are you thinking you will go commercial with it?
Congrats!! I know you have been beavering away at it for quite a while. Wish my rig supported the newer protocols so I had something to test and develop with! I did purchase a CAN module to write software with but haven't found the time yet.

"do things like treat a shade as a shade instead of as a light bulb." myServer (and many other automation systems) treat a shade as a lighting device. The "dim" value opens the shade to "50%" open. Shades fall into myServer's "lighting family" (a class of devices).

I would be very surprised if Newmar took interest in it. I suspect you will have to go direct to consumers. Would be good to start a list of "known compatible with" and test results.

 

[flyboy013]

Hi flyboy013 - I was wondering if you made headway.

I’ve made quite a bit of progress. I currently have integrated with both the RV-C and KIB networks. I’ve been improving things as issues or lack of capabilities pop up. At this point in time, I think I have a very very good beta capability. Still finalizing the HW.

Didn't see you post on that "other" forum.

I had posted an update on the "other" forum a while back, but the moderators had issues with it. Whatever.

So, tell us a bit more about your KIB adventure. Did you find protocol docs? Squeeze it out of KIB? Reverse engineer it?

I did not find any additional KIB documents. I (with some help) was able to analyze the KIB capability and integrate with it. I won’t say that it was reverse engineered it, as some may have issues with that terminology.

Are you talking to KIB via the RV-C network? Do they follow RV-C standards or did they mod it a bit to make it proprietary?

The KIB network is not a RV-C based network.

Does your hardware talk only RV-C? or other networks as well?

It talks to both the RV-C and KIB networks. It also talks to other components (ie. electric waste vales) via MQTT.

Does / Can your solution talk MQTT by chance to integrate with our myServer automation system? If so, we could very quickly add support for it. If not MQTT - any other integration methods supported? Are you thinking you will go commercial with it?

The solution does utilize MQTT, but it is not designed to be an interface to external systems. The MQTT server is currently userid/password protected. The goal is to secure it further with SSL, limiting access to LAN, and restricting accounts to reading/writing only a portion of the data in the database. I worked developing DoD systems for quite some time and understand the need for securing the system.

Congrats!! I know you have been beavering away at it for quite a while. Wish my rig supported the newer protocols so I had something to test and develop with! I did purchase a CAN module to write software with but haven't found the time yet.

"do things like treat a shade as a shade instead of as a light bulb." myServer (and many other automation systems) treat a shade as a lighting device. The "dim" value opens the shade to "50%" open. Shades fall into myServer's "lighting family" (a class of devices).

I total agree that many systems are treating a shade as a light bulb. That was I original didm but I didn't like having to tell Alexa to turn a shade on/off in order to raise/lower it. It just doesn't make sense. That's why I decided to develop an Alexa skill. So that I have just better control of the various device. My goal is to be able to ask "Alexa, what is the house battery level" or "What is the grey tank level" and have it tell me the appropriate value. You cannot do that by treating devices as light bulbs.

I would be very surprised if Newmar took interest in it. I suspect you will have to go direct to consumers. Would be good to start a list of "known compatible with" and test results.

I suspect you are right about Newmar not being interested. I'd believe there is a market for this capability. I wish I had already push this capability out. It was my intent to have it done, but to be honest the KIB integration ended up being an issue for me that just took time to resolve. Now, my biggest obstacle is designing a PCB. It's a new technology for me, but fortunately I have a few kids that have provide me needed guidance. Perhaps the biggest issue is dedicating the necessary time while I'm not on the road. Seems like when I'm home there just too many things that need to be taken care of!.

 

[AbdRahim]

Looks like great work. I guess I am old school. One of the reasons I went with the Bay Star as opposed to the Tiffin open Road was switches - no computer panel. I can fix switches. Where are the wires my panel is controlling?

Hats off to those who love that stuff.

 

[DavidL]

"The solution does utilize MQTT, but it is not designed to be an interface to external systems. The MQTT server is currently userid/password protected. The goal is to secure it further with SSL, limiting access to LAN, and restricting accounts to reading/writing only a portion of the data in the database. I worked developing DoD systems for quite some time and understand the need for securing the system."

Are you saying that with proper credentials, that a third party system couldn't use the MQTT to bidirectionally control KIB / RV-C devices using your solution?
I would rather leverage your solution (including pointing people to buy yours or resell yours) than spend the effort to reverse the KIB protocol for the small number of expected people that would want to use it (relatively speaking). The two systems would be on same LAN. The user IDs would belong to the owner of the RV. Certainly we would need to know how to pass and accept the authentication method you use.

So hardware wise, on your RV - is your hardware on a breadboard? And your talented kids are helping you get it onto a PCB?
FYI, there are lots of free PCB design software (Kicad is one). Once you have your circuit in the software, you just email the resultant file to one of thousands of inexpensive PCB manufacturers that support that output format. Then a week later, you will get the PCB populated or not (depending on what you choose). Easy peasy.

Would be nice if the PCB was designed as a Raspberry Pi4 Hat.
Another thing to consider is I assume KIB is just a CAN based protocol. There are lots of PCB / Hats for CAN readily available. Just write your software for it.