Reality Editor proof of concept video

I have been working with some support from Audi and Steelcase on a proof of concept for exploring the capabilities of open hybrid and the reality editor.

You can find the project here:
http://www.realityeditor.org

What do you think?

2 Likes

@valentin this is simply beautiful :smile:
how much of the functionalities are implemented now?

Do you refer to the video?
Everything you see is implemented and fully working.

We got some serious support from Audi. Beside the car to experiment with,
we had engineers from Germany flying over and woking with us to implement all the functionality.

2 Likes

@valentin i was referring to the Solid talk video in home page ,
This is an amazing research :smile: , is the car and iphone connected to the same private network and where does the server run ?

There is a openhybrid server in the Car and everything is connected to a private network. Yes.
I don’t know how many Audi cars are out there that can be controlled with an Arduino. :wink:

1 Like

Very cool. But you have a typo in the link :wink:

ups, not anymore :blush:

I love the PoC video and the vision it presents - this interface is exciting.
I stumbled across the video over lunch today, that got me digging for more info and I fortunately ended up here.

I don’t have a YUN (yet?), but I plan to give this a spin on my Pi over the weekend @V_Mohammed_Ibrahim I saw that you got the server/ui up and running, that’s awesome. I might ping you if I run into issues. Any success on interfacing with the GPIO ports?

2 Likes

Hello Kevin,
Welcome to the Open Hybrid Community!

Greetings, Valentin & thanks for the warm welcome.

Greetings all. I happened up on the video and ended up ordering a Pi to play with - this is great stuff! It’s been a few years since I’ve programmed (I just manage people now), but I hope to pick up javascript and get something working by following in the footsteps of those here. I have lofty visions for where this could lead in my business space, both practically and as a stellar marketing tool. Heck, I’m sorely tempted to point my team at this and hopelessly distract them from paying work!

I do have a couple of newbie questions, if you have the time. I get the notion of free-standing IoT devices that manage their own presentation layer, but for less capable IoT devices wouldn’t it make sense for a local server to proxy for them with a less demanding low-layer protocol (GATT?) to the device? Why not just use a PC server instead of Arduino/Pi or such – wouldn’t anything with node.js plus appropriate interfaces glue things together?

Is there an outline of where this architecture is intended to lead, in terms of scalability, security, and other industrialization needs? Or is it reality editing so new that this is bleeding-edge proof-of-concept, and the future is evolving over beers and espressos?

Regards,
Peter

I have some similar questions. I was going to bring them up elsewhere, perhaps in the diagram thread, but I’ll post here to keep things together.

In some older demo videos, it appears that OpenHybrid (or one of its ancestors) was AllJoyn based. Why was this architecture abandoned?

Does OpenHybrid (plan to) handle discovery and communication with devices that are on different networks (but still internet connected)? If so, what does the security and authentication story look like?

Personally, I’d love to see this up and running on an ESP8266. From what I see, that should be possible - the discovery and UI layers won’t be a problem at all, and the only challenging part about the programming layer will be reversing the current implementation and creating a new implementation in C. Any thoughts on this would be appreciated.

Yes, you can absolutely do that. All you have to do is create a hardware interface (there is a plugin system in place for that) which can somehow communicate with your IoT device. That’s basically what I did with the phililpsHue hardware interface. I created a hardware interface which can communicate with the hue bridge’s REST interface (little warning here, you don’t really want to use that hardware interface, it was mainly created to test the new hardware interfaces API I was working on, so it’s got no error handling and not a lot of functionality yet). It really doesn’t matter how you communicate with your IoT device be it bluetooth/GATT, a REST service or whatever. As long as your “HybridObject server” can communicate with your IoT device you can create a hardware interface for it.

2 Likes

Hi @psomdecerff Welcome to the community!

These are all very good questions.

I believe that in 2-3 years the processing power of your less capable IoT devices will have reached the need for the Open Hybrid platform. I find it relevant to rethink the entire IoT concept from the perspective of objects that know everything about them self.

It can scale infinite, because it runs in private networks that are completely isolated from one and another. In each private network you can have up to 1000 Hybrid Objects. The decentralized aspect is key. There is some beginning research from IBM in using Blockchain technology and Peer to Peer networking for connecting your Objects via the untrusted Internet. I think this are the right kind of explorations. I will try to advertise this a bit more and hopefully get some people from the Bitcoin communtiy on board with this. :sunny:

At this point, your router is the gate keeper to your private network.
However, since everything is build with node.js and standard protocols you should be able to use https for your GUI interfaces and a login. In the long run, I would love to discuss the options that come with Blockchain technology a bit more. But I am not an expert on this. Maybe we can find a group of people who would like to explore these options?

What do you have in mind?

Kind of :smiley:
We need to discuss and solve many of these questions. Do you want to join our OH Conference for more of these discussions?

2 Likes

My thinking was the following: If I find a hybrid object in 50 years on my attic, how would the technology need to be implemented so that I would have no problem to power it up and use it with my infrastructure.

I was looking for the most stable and most known foundation that will serve this stable future object.
I felt that REST, Web Sockets and JSON would do the data communication job perfectly.
These are open, established and a foundation of the internet.

At the time I had the impression that AllJoyn tries to reinvent the already well functioning wheel.
So I decided to implement the most common technology instead.

I envision that your questions will be solved using Blockchain technology. IBM has made a proof of concept that private objects can communicate decentralized via the Internet. We should discus these questions at the next OH Conference

For now, think about Open Hybrid as a platform for experimenting with connected devices and Augmented Reality in your private network.

At the last OH conference we had somebody who wanted to work on that. I will see if I can get him on the forum. Originally I started to implement Open Hybrid for the TI cc3000 + Atmega but it took ages to get it right. I felt that a higher level system would work much better, since in a view years these SoCs will have the same price tag then the atmega.

felt that the common open internet standards already solve our problems and they

@GeorgeHahn
You could include external services such as IFTTT via new IO-Points or dataPointInterfaces.
You can find the default example for a dataPointInterface in the object folder dataPointInterfaces.

The idea would be, that your object can have a “internet” service data point that you as a user would need to intentionally draw a line too. As a design guideline, the Hybrid Object data points should never be directly linked to an internet service, so that a user can intentionally chose when to leave the private network.

But all in all, the visual interface is not fully capable of such interactions at this point.
I plan to solve many of such GUI problems with my PhD dissertation.

Thank you much for the prompt and thorough response.

I find myself in the enviable role of providing technology for a “late adopter” market, namely energy and process automation industries. I long ago determined that my value is largely in getting nifty new services to run on obsolete technology and in somehow keeping obsolete services running over new technology. So, I will confidently state that at some point I will have 15-year deployed legacy equipment on recent deployed RS-485 networks alongside brand-new wireless Hart equipment tied into a redeployed host of some sort, all with expectations of managing them via open hybrid. By the time I universally have broadband networks hooked to high-capability IP equipment, you all will have moved on to something else entirely (or had grandchildren, retired, and so forth).

So I agree with the vision, but add a ball and chain of backward compatibility to it.

As for self-description, to paraphrase Einstein, why would I need to know something about myself if I (and others) can go look it up?

Ideally, where you see self-describing end-points, I will also see an interop gateway with associated descriptions and a legacy device. This seems perfectly fine to me - the “modern” IoT gateway will be like you say, only with a bunch of legacy tendrils out to obsolete devices. I’ll build a gateway, and it’ll support Modbus, Hart, RS-232, RS-485, analog and digital IO, etc. on one side, and REST/OH on the other. A casual user shouldn’t know or care what the physical media is.

Device discovery and acquisition (I can add a new endpoint to my network, but you can’t take one of mine over), easy failed device replacement. Zero-touch network add and one-touch initial configuration (a perpetual pain on shared-media networks, like wireless). Compartmentalize and disambiguate if there are 1000s of similar devices in a plant. Role-based views and restrictions (I can only see current data, but you can change a config). Context-sensitive help and context-assisted operations (pull up manufacturers webpage user doc, or shell into my local GUI to do something complex).

Existing specs (like DTM and FDI specs for Fieldbus and Hart networks) have a notion of standardized presentation with a modular approach.

I would much like to. I doubt I will get company support for this, but depending on where things go over the next month or two I may join anyway.

That is Stunning…!