Mario Hoffmann - Roadmap to Life Management Platforms

Start and let Mario give his presentation about life management platforms from a research point of view and welcome Mario. And thanks for participating. Go ahead and get started.

Yeah, Sure. And thank you for taking me first, so I can just consume all the time for first hour.

So yes, good afternoon. I am Mary Hoffman from F Hofer research here in Munich's the Institute for applied and integrated security. So today I will just continue actually with my presentation from yesterday, there was already a little session on life management platforms and life management initiative. So what I would like to share with you today is a little bit more on the research side. So yesterday we discussed already a lot on use cases, the business models, standards, and so forth, and what could be part actually of a life management platform. What are the requirements?

And actually the requirements is the first, what I would like to share with you. It's just, let's say, yeah, actually it is the last slide that I presented yesterday. And I'd just like to make sure that we have the same understanding in terms of the requirements when it comes to life management platforms. So the first one is, and there was already a lot of talk about privacy by design. So privacy built in privacy is a very important requirement.

I think for life management platform, as it is supposed to deal with very personal, intimate data that yet you can control and you can manage and you can especially manage what is the purpose of this information and where does it go to? So the next big thing here is information storage and sharing.

Of course, these are fundamental functionalities of a life management platform. No doubt about it. Important here is interoperability. So yesterday again, we made the experience that it tends to be actually to have, again, those silos though. They are some approaches from the telco companies, yes, life management platform, very important. Some from maybe cloud companies, you have personal cloud, this is actually life management platform and you have already established the Facebooks, the Microsoft, the apples, and so forth the Googles.

And I think they would consider themselves as well as a life management platform. So interoperability means here that you can export your data and just take it and put it into another supplier and another supporter of life management platforms. So the generic life cycle management, of course, in a life management platform, you have a lot of identities, profiles and so forth. You would not only like to create them, maintain them, change them, but also delete them. And this is one of the other major requirements for a life management platform.

And it refers to the proposals and the new data protection in you, the right to be forgotten. Well, Don't forget the existing identity providers. So we have to integrate them somehow in such a live mentioned platform. So there's some kind of old world that we have to take into account.

Of course, if we design and specify what a live management platform is supposed to be so authentic user data is of course, one of The good things here because for a user life management platform means that there is one central 0.1 dashboard. If you like where you can manage all your relationships to requesters relying parties. And here again, we come back to the integration of the existing identity providers, but at the same time, you have this storage functionality. So you can fill personal data also in this life management platform.

And this is very authentic because it's the only place you have to manage it. I mean, typical cases, if you move, for example, you have to change your address. And then you have to remember all the relying parties where you have to update this information. And here you have the chance just to have one central point to change it once. And then it's propagated to the other parties that you have been in contact with Mario. Wouldn't it be safe to say that if you have a life management platform that you are, have become a service provider?

Yes, of course. Yeah. You're now a service provider. I will become a service provider.

Yeah, That's right. And you have APIs that give you access to your services. Yes. And they need to manage the ingress and egress of data from your yeah. Service.

Yeah, That's true. So transparent user control here is also an essential thing. User empowerment. One of the big buzzwords here, I mean, forgive me, those are very high level requirements. This is true. But we are just about in the face to design and to specify what life management platforms are supposed to be authorized access and policy management.

Of course, if you have such a central point where you can manage your data, it's also important to specify, to define policies, authorization rules, who can get access to what and when, for how long and so forth. And somehow this is all related to and sorry for this password.

Bingo, big data. Very cool.

So, but big data is just a topic for another session, I guess. So what I would like to share with you today here, because I'm little bit also a researcher, so it's kind of well, paradigm shift, well, sounds pretty large, but in case of what I would like to share with you, it's kind of a change in the transaction model. What I would like to suppose and propose to you is just assume, just imagine you are an app and this is actually highly related to what you just said. The user is a service provider. So if you are an app, you can be certainly downloaded, But downloaded by what?

Well, for example, you can be downloaded by your hotel room. So just imagine you have maybe a link on a little NFC tech or your phone and you enter your hotel room and there is an NFC reader and you hold your device against it. And then the hotel room personalizes. So wouldn't it be cool to have the lightning, but also what maybe more interesting is your favorite TV stations in the same order, the same sequence that you have at home or on some kind of picture frames, you have some of your relatives, your wife, maybe your favorite football game, your football club already showed there.

So in this case of personalization, this room needs some kind of information, of course, but where does this information comes from from the life management platform at the end, The life app that I would like to introduce here, maybe you have seen the animation, which goes from here, the life management platform down to here. That means the hotel room downloads you. This supports actually a secure channel because now you have here a little trust anchor, well, this is your piece of little software. Maybe it is already, it contains already just a link to your life mentioned platform.

If the room requests more information, maybe it's already, it already contains some master data, maybe some preferences, some policies how to use your data, also the usage. So the usage, how you use the room could go back feedback here into your life mentioned platform. If you are in favor of maintaining histories, for example, of how you use hotel rooms in the past. So here are some examples of what you could personalize nowadays, lightning floor heating and so forth.

I mentioned it already, But of course not only those hotel rooms, I mean, we can generalize a little bit, so you could personalize contextual environments. So also for example, vehicles, so cars in the car sharing concept is very interesting for this one, but also all the online services that you have already, but also other people maybe in general, what if you go on protocol level for the general aspects, just referred to referred as relying parties and requesters.

So here you have little pictures, just little examples of other contextual environments could be vehicles, as I said, but also a supermarket, which just leads you through the floors to get the food you need, the office environment, or of course, ambient assisted living. So eHealth was also one of the big use cases, applications in ours that we discussed briefly yesterday. Yeah. And let's call this app life app. I think that's a pretty obvious name for such a thing that is directly connected to your life management platform. So here's just some summaries.

What I've tried to illustrate you with those little figures. So what I introduced already is if you have different users that share their life apps. So this is kind of an example, which is pretty good here for this kind of conference. So we are just together for a week. And after that, everybody goes home back to the offices daily life. But what we share here could be important just for this week. So let's be some kind of buddies just for the week.

So if you go downtown Munich and see an interesting restaurant, you would like to share with it, share it, this information, maybe just with the people you have met here and not with your other 500 plus Facebook friends. Cause for those, this information might be not that interesting, but just for this context here, this information might be a good one to, to exchange. We have other examples with things, what are said with your rental or shared car personalize your car, you have your favorite radio stations.

Maybe you have your, your, your history of the places you have been to just for the next destinations. You have your smart hotel room. I illustrated this already, maybe friends that visit your, your own home, your own house could also take advantage of your intelligence in your smart home. And of course services.

I mean, currently we have lots of transactions interactions with the services. So here also in e-business processes, this life app could be beneficial. We have a second. For example, in the eHealth domain, nursing service has access through a life app enabled monitoring service provided by the users, body sensors. So those are, let's say a little bit future use cases.

Of course, let's say three to five years from now, but that's actually what we are currently trying to cover in EU projects in nationally funded projects, but also in contract research. So we also thought about a little, how could this look like? So yesterday we had this little presentation from the key people. Maybe they are a little bit ahead.

I mean, I'm not an expert in user interfaces, as you might realize, but what is important to me is what kind of functionality should such a life mentioned platform offer? So we have here on the left side. Yeah. Creating. So creating means virtual identities, profiles, preferences, policies, whatever new relationships I can, of course edit them. I can share information. So here in this little window here, we have the registered peers that I have some sharing relationship with. So above there some users, then we have some environments and here we have the services.

Of course I can close those relationships and I can follow specific ones. So in all of those cases, this, this kind of transparency aspect is very important. The second aspect, which is very important is empowerment. So you can create new virtual identities. You can be sure that the subscribers of your personal profiles taking advantage of the life app are up to date. As soon as you added something here in this profile, because it's just propagated. Whether it's a pull mechanism or push mechanism, I don't care at the moment. This is just then the question for further implementations.

But you, and this is the most important thing here. I would like to see the user in the driver's seat. So you can control it anytime who has access to what and how long, and you master the identity lifecycle, which especially especially includes that you can terminate your virtual profiles and relationships.

You some, just some other impressions where we just played around a little bit with the user interface. Yeah. It's pretty soon and pretty fast. It's get very crowded. So we need to find some other and better usability metaphors, I think for, for those kinds of management tools. But this is actually the information that immediately comes into your mind. If you think about managing your life now, there are some other ones. If you'd like, you can just have looked at in the slides after afterwards, this will be available.

Of course, A little excursion about the things that we are currently are investigating, having underdevelopment. So this is what I call, let's say a hello world, example of a life management platform.

Of course, this is a little bit more complicated and complex like than the, the other hello world example that you know, from your programming times. So what I'd like to do here is just to get some kind of newsletter service. So here on the right side, you have little relying party which offers service and here with your web interface, you just go to this service. And I mean, without going step by step, all the 11 steps here in this session, the main important thing is that here on the right side, this service provider has already downloaded your life app.

So the thing here is, and this is what it makes just a hello world example, this list auto fill in functionality. So you go here, have this newsletter service, the newsletter service is asking for some information, but you don't type it in by hand. You just click the button because the relying party is already LMP life management platform enabled. So it goes like this here and here you have, then those typical redirections. And in this sense here, the life app goes to your life management platform gets the information.

Then there is supposed to be some user and the form will be filled out automatically. Another one.

Yes, my, My guess is that how it'll really turn out is that that relying party won't download anything except the data that comes from the API call they make. In other words, you give them the location of your API and the token to be able to access that data. And they go get it in realtime rather than download any application or data sets or anything like that. Yeah. Well this is research a proposal, right? You're right now nowaday nowadays, I would also go for the first step that we don't download an app, right.

Or, or, because this remembers me also this good old times about this multi-agent systems that failed from 12, 15 years ago. So I know all those difficulties with that, but I also see the advantages if you have this kind of life app at the service provider side, and the difference between this approach and the multi-agent systems is that now we have those mechanisms established.

So we, as a user are used to download apps every day. So why not just shifting the whole thing and convince the service providers download me and you have all the information you need Because see, I out update too fast and the API doesn't update, it gets updated in real time. So I'd much rather standardize how the API works than standardized the practice of downloading my app. Yeah. And just make a call to the API, get the data in real time. And when we're done.

So to Me, it looks like, you know, this app can actually be represented as the URL, The restful Make, Make a restful call and authenticate with, with O get the data and the session. And in many cases that subs provide down below actually these be kind device, but it's, It'll be all over, But I don't, this is a contradiction To no, not at all. I'm just suggesting It's implementation to be populated through the, but it's very nice and good to have a visible thing locally on your device where you can easily handle browser.

Things will appear strongly Good by we're all in agreement. I think possibilities. Yes. So what would, what would be the advantage being an app that just APA? I can't see one, I have the, on the last slide I have just the comparison of between several benefits. If it's not in your direction, maybe we can take the question then. So just another example from our current research, this is a little more complex scenario. Just assume you go to your website of the town hall. You are a citizen I would like to apply for social benefits.

And then, then it's not only filling out one form with little data that you have in mind already. This means that you have a lot of different information that you have to fill in, in several forms. And maybe not once this is a process that you have to come back to those websites again.

And again, if you do it today, it's like you get this kind of paper. You go home, try to fill it in, try to get all those information that is needed. Like bank accounts, patients records, and such things that they're asking for. Then you go back, of course you have missed something. Then you go back home, look for the next. And this is four or five, six times. This happens until the process is done. But with the digitalization of those citizen-centric services, I think this can be really speed up. So what we need here again, is this kind of, yeah.

A little bit more complex life management platform that we have seen on the slide before. I don't go through all those boxes now. So don't worry, but important is of course the red one, the personal data store, which is one central element that we are currently developing. It's not rocket signs, but it has to be done just to make all those pieces work together. At the end, you need some kind of authorization manager where you can manage your policies and so forth. And if you think that this remembers you somehow this kind of green circle here.

Yes, but this is just Canara Yuma. This is a little adaption of the protocol that has been these specified last year in the contour user managed access group, which is actually based on all 2.0, of course we have this opener Deconnect things also implemented. So in the current work, of course we are based on the standard protocols. So what do I see currently in terms of benefits here for end users? It's just like a little summary. So I think I've mentioned platforms, of course, can empower the users to maintain all of their personal data with only one central management tool.

And this management tool has the chance there's some personal data store, but also some kind of possibilities to maintain just references to other data sources that you have already users have the full control over the service, the identity, the information lifecycle, and here, of course, it's possible how we are going to implement this kind of policy management with downloading the user's life app. This is kind of our proposal. Then the relying party has to agree now on the user's terms and conditions, this is again a specific, specific thing here for the life app.

It just turns the world around because nowadays you download your apps for five, 10. I don't know how many you have on your phones. And usually you have to click, okay. I accept the live, the terms and conditions, and usually you don't read them. So why don't force the service providers agree to your terms and conditions by downloading your app? Because they would like to have your personal data, isn't it? That has key consequences.

For example, in the way, service providers are able to handle their clients Again, please conditions Are, are not standardized in some way, you know, some boxes I would allow you to access certain type of profile level of profile level of access. If it is individual might become close, impossible to deliver services to an individual, isn't it? Yeah. Yes. This is true.

Well, I mean the easy case might be, if you just, or if you make the service providers agree on EU data protection law, or if it's kind of a company just as, as to agree on your company's policies, security policies and so forth, this could be standardized. I agree if you go on the individual individual level, sorry, it's almost impossible to have a very standardized way to do it. So just as an information back in Japan, we are talking about like that, how to reverse just like this, you know, the conditions or allow person license there, there for that.

We need to have pretty simple set of licenses, like creating like that. Otherwise it's not apparently full is also something heading towards Also possibility for user to earn some money. Because I can imagine that you can have a field with, for example, a political opinion or intention to vote. And on the other side, somebody would pay for that information. Yeah.

Well, I would say I'm not an expert in defining business models. So as a researcher, I'm on the technical side and sometimes, well, actually I'm not in favor of service providers, which business model relies on selling personal data. Okay. Here it's the other way around the user can make some money out of it and can participate when those relying parties make some money out of it. Okay. I would like to have my share.

Yes, yes. Maybe, Maybe doesn't.

I mean, if, if we, maybe if you ask for the rule over here, then we sell this data. I would not raise my Head.

No, I wouldn't need it depends on the use. There's all sorts of reasons. The attention the people give of information for free these days in, when they go on the give free, We talking, which so in such an environment I would probably, or even nowadays environment, I would, I would not see a distance focused on the fact that I, personal information Agreed Maybe as an add on, if I do something, I get a better service. If I give a bit more of my personal information and agree into one certain research, like for example, I get a film for free.

I would like to buy, I can choose either I pay four euros or I that's example, there might be something or something behind it, but it's the same stuff. If you get fuel for free or you get five euros and you buy fuel, whatever you want hotel. So for something that like that what's important is whether the user actually is understanding the sequence of doing that's very D actually yes. That's about meaningful consent. Yes. I'm already a little bit over time. So It's up to you. Sorry.

No, it's up to, up to the writer. We Have extra time. Go Ahead. Okay. So Question down there.

Oh, I'm sorry. Over there. Okay. Sorry. Going hopefully there in the end right now we have, for example, companies, data selling personal information. So they're our personal information. So for example, you're the control of your data and you're able to hundred and you know, what's information, basically the ownership. And so I few think it's not vital. Why not? Or why?

Well, the reason I wouldn't do it is I'm not, I don't want to be in that business and means I gotta spend my time managing how I'm going to collect the money from people using my personal data. I, I, I'm not in that business. I don't care about that. I it's money on the table. I just assume we rather than spend time managing it. It seems to me that those things can be pretty much automated to save It. Can't it's gonna require your attention one way or another your brain.

Can't, you know, I, it's also a question and information portfolio, portfolio companies, you know, who, who did that? You can earn a lot of money, like three, Your honor. I'm told if I clip coupons, I can save at the store and I don't do it because I don't want to clip coupons. So yeah.

I mean, it's just up to you, whether you want to be engaged in trying to make that as a money income opportunity. Hmm. And I understand you when you say you don't want to be in this business because if those people don't pay, oh Jesus. So this can be really nightmare.

Well, I, you know, I could, I think that I can make a lot more money being, thinking about 'em being engaged. What I love doing rather than what would be a mundane process that could, didn't take a lot of my time.

Like, I don't want to do that. Maybe somebody else will, I don't care.

I just, I don't. But I mean, this is the reason why, why we are here. Right? So discuss the life management platform stuff. Why should be Personal?

You know, it's up to you, It's your data. Exactly. Figure out what you wanna do with it, but don't make me do it. Okay.

So, yeah. And, and the last benefit I see here from the user's point of view is this kind of trusted anchor. If you have the relying party downloaded your life app, so you have some kind of anchor on the service provider side, you can establish a secure channel with this one and can control it. And you have then a little piece in the infrastructure of the relying party that is owned by you. But I see also some benefits here from the service provider side. So there is then a single point of contact with this life management platform.

That means that service providers, whether they come from people, things and services in the future internet, I don't care. But the thing is that there's one point where those services can contact me and can ask for privileges, can ask for information, can ask for whatever data they would think they need. And I can then authorize this or not. The other point I already mentioned is that those subscribers in this people, things and services, internet, they have access to up to date and authentic data.

Assuming that a user is really willing and motivated to maintain this data, because it's only one point where you have to maintain it. So not 10 20, 30 different service providers, where you have some information here and here and here, no there's one management dashboard where you can integrate everything and manage everything at one point.

And what we have discussed a little bit though, this is I see a little advantage also on the service provider side, if it comes here that the service provider has to agree on the user's terms and conditions, they don't have to take care of, especially if you are a global service provider, what kind of local compliance to data protection, for example, do you have to fulfill, to interact with this user? So this could be something that come, could just come with your life app, either as a link to those regulations or already built in. So those are the things that I see as the benefits.

And now I can come back to your question. Would you like to ask again?

Oh, So my question was, what's the benefit of being app rather than API? Well, the app, as far as I understand the difference between API and those app. So with an app, I mean, there's some kind of code that the relying party has to run their side. So in this code, there could be already wrapped some kind of master data, some kind of policies. You could have your policy decision point and enforcement point at that point.

And you could decide already some simple things, maybe at that place, without going back to the life management platform, asking for more things and having kind of protocols going through the internet again and again, there could be something already, some pieces that are built in, in this app, maybe some default process that you have built in, and then you can interact directly on the relying party side. And I put it to you that you want the agreement metadata to happen every time. And that's because it's guaranteed that whatever they downloaded last time is outdated.

And so if they're, depending on what they got last time as the relationship between you, it's probably gonna be wrong. And so you, you know, I mean, even this is how DNS works.

I don't, I don't go do a name to address resolution with a cash. I go do it right now. So I get the right one. And that's what you want with life management data, I think is I want to hit the API. If there's metadata that gets downloaded, that deals with the service level level agreement. It's current.

It's not, it isn't outdated. It's not stored somewhere where the relying party could make a mistake.

I want it, I want it to happen. Now. That's why I think the API approach is gonna be much more valuable in a life management platform scenario than having to download code that gets run. Yeah.

I, I see there are some issues with this app approach. I I know, and of course going for the API approach first is obvious because it's easier implement, You know, it was complicated and it wasn't so easy. So it wasn't so transparent that that's the way you should be doing it now. And I think that we've simplified it to a point where it, it doesn't make sense to make it happen, to be an app in my Mind. Yeah.

I mean, what you could implement in an app of course is little synchronization. So if you are worried about the out data data that might come or might be placed in, in this little app, that could be a little request, Hey, life mentioned platform, is there really an update happen? Then the relying part has to agree to that update and you know, it's friction, you know, you wanna reduce that anyway. Yeah. Their choices. I think the API journal better.

I, I was thinking that they just take the same mechanisms the app providers use today. So there's always something happening on my mobile that I don't know, some update mechanisms synchronization with 10 apps and parallel. So that's why I thought, well, it could be also. Yeah.

But if, if the apps updated, you have to go through a process of saying, yeah, I'll, That's take that. You know, I don't do that. Okay.

Oh, so there's another question perhaps More and more follow the, I am fully with you, but on the other hand, I find the application, the light app by the traffic, but because how do I show the room that it's me, that I want that, that room to be configured. Like, I don't want to go always around with my PC.

Perhaps I want to use a QR tag just because it's, it's just inexpensive to be scanned whenever where I am, perhaps it's NFC then like that maybe part was probably a good thing do, because my phone I have with me all the time, and the other thing is also these, this, now I have this room that is requesting well ion now requesting I want to have access to your data in order to know how should I dim the lights? What type of music do you want to listen to?

Blah, blah, blah. How do I confirm that this is okay again, I have my phone with me or I have my phone with me. So the live app popping up, DIA wants to know how to configure my room. Okay.

This is, thank you. Maybe one thing to add, I can imagine situations where I would explicitly like, and then to run online hardware, compare insurances of different provider. The business model for the app is selling the app to me, guaranteeing that it is independent from certain assurance vendor. This is a nice business. I like this idea. And if such an app can run in my personal feelable ratable environment making, you know, I can, I can hold it. Sure. Nobody else will get information that I don't get.

I think this Wondering whether this is compatible with the trend into cloud, basically having all, a lot of information gave From the cloud Devices are devices basically just displayed, right? So they're not displays. They're more like directories, right? So I'm gonna, I can have the, the basic DNS data here that lets me go where I already know the data's actually stored in the cloud. Or I can go take a look in the cloud and find out the rest of the metadata that lets me get to it. But what I don't want to do is prescribe where the data is or its usage.

I simply want to provide simple access that then gets based on the event that occurred, that then has a policy attached to it. That then does the action. So I provide an event followed by an action that is controlled by a policy. And that way everything remains loosely. Coupled I can, I can manage how it works in space and time with low friction. I wanna be able to do it as things change all the time. If take a different persona, have a different scenario. Can my schedule change today?

You know, I don't want to wait for that to ripple through a database that's been prescribed that has to then download the application and get permissions to do it. You know, it's not gonna happen. I won't keep up.

So, you know, in the life management platform, I want it completely separated. I want it to happen in as, just in time as possible. And I want it to be in the cloud so that it, so that it's location, doesn't matter.

You know, if that, if that server goes down, guess what the replicated one picks it up and, and it's available and runs all the time. Anyway, you know, that's the, that's the kind of persistence and availability that I want no matter what. And that's why the cloud makes a lot of sense. So Expression, the only method of deployment much More, right? It's not the quality of deployment. Yeah. The only thing that I would add to what Mario said is the definition of interoperability.

You know, his definition of interoperability is that I can take my life management platform and go move it to another cap, you know, to another side of it. And my definition of interoperability is that I'm not committed to any silos, but that the interfaces between the functionality of the like management platform are so well defined that I can have my choice of which service provider provides that kind of service. So my personal data store, isn't a silo. It's simply a service that I've plugged in and chosen to and agreed to the terms of usage.

And if I'm unhappy with it, I jerk it out, put not a different way. That doesn't mean I have to move all my data over to some other silo. That's that's the, the big difference I would add to the explanation of your definition of what interoperability means. Yeah.

Well, I propose that we take a coffee break for 10 minutes and then come back and we'll do the next session. Maybe I, sorry. I have another slide If you'll hold on to that thought for a moment. We'll see. Mario's last slide.

Yeah, this would be great. I thought you were done.

Sorry, Because there's just, there's one question we have already tackled and covered a lot of different aspects and questions, but maybe what we haven't is who's going to host such a life management platform. Well, I, it's not gonna be a silo for me. That's for Sure. Yeah.

And I mean, yesterday, we also had some discussions, could it be a foundation, like key, could it government services, maybe this is something special that we have to have to, I, what I have learned from the Scandinavian people that they really trust in, in those EC government services though, for them, it could be a solution companies, banks, telcos, internet gang, whoever, what I like a little bit is this kind of you? So I have my server at home, but maybe this is a bit, yeah. I Want my own domain in my own cloud and that's where I put it. Yeah.

And I pick and choose for where all my other data comes from. What I also can think of is data protection authorities. Cause they are the last people I assume who are interested in selling my, my personal data. And as soon as there's something behind it, and this will speed up a little bit, probably all of them will try. So that's actually my last comment and a little conclusion here is yes.

I mean, important is that we try to get in common understanding what life management platform is supposed to be and how to integrate this in the future internet of not only those online services that we have already talked about, but also with people and things. And I would like to encourage you to join the life management initiative that has been discussed yesterday. It's not funded yet, but it's that close. So we just keep you informed about those activities afterwards. So that's what was thanks For my own. See you back in 10 minutes.