[00:00:06]
>> Glad to see there are some faces in the room that people could make it come in so again as were not mentioned at the beginning my name is Shannon McCain I'm with the University of North Carolina and I'm the new deputy director of the south big data hub so like some of you on learning a lot and hopefully I can learn something from those of you that have been involved for a long time this next panel in presentations about the spokes some of the activities the research that's been happening under the Hobbs umbrella and S.F. over the last 4 years so we have 6 speakers and we're fed up with similar format to the earlier panel where you speak will come up and talk about his or her research and then at the end we'll bring them all up and you can ask them questions the one exception being the 1st gentleman Gary Clifford who's going to talk about a large scale medical informatics for patient care coordination engagement from Emory University he has another gauge MIT So if you have questions for him make sure you need to ask him during his presentation Gerri thank you.

[00:01:08]
It's remission can you think it all right. I be back in the afternoon as well so I mean you're going between and Georgia Tech which is a good metaphor for the fact that I also have a position at Georgia Tech and. More houses around so. That it's a good analogy a good representation of what we're doing in Atlanta so Rose working together it was working across the board as and trying to leverage the different strengths of the institution so I'm really proud to be part of different months.

[00:01:43]
Let's see. Here is my son it's so. This is appearing on the ground with Doc to have intelligence that the director of the kind of best to research and shoot at Morehouse School of Medicine and he spent money the director of the Jackson Heart Study. And I've been very lucky to work with him over the last 3 or 4 years and also read Dr Bob you know who is.

[00:02:13]
One of the copy eyes on the grants. From Texas and we've been working together on this for about 2 and a half years now and we're starting to see the fruits of I write this I'm just going to talk very quickly through some of that and then just open it up to questions.

[00:02:31]
OK so the over here is that heart disease is is an enormous killer as we know. But it's as you can see from the disproportionately affecting this in this strike back region. Of the southeastern United States is getting dressed in the US over time in terms of its proportion of bad on heart disease.

[00:02:54]
Rather think it's particularly difficult with heart disease of course is that it's a many time problem and it's curative behaviors and cumulative experiments so let me 1st turn to my doing this project that I came to me and said I'd like to do something similar to the hop the Jets the hostility that I write.

[00:03:15]
In a mad my 3 cell phones and try and do a ubiquitous way of monitoring of people's behavior and expenditure is over time to see its effect on cardiovascular disease but just she's in the devices Abreu already have in our pockets and in the right cell phones or any other.

[00:03:35]
Type devices that we have so many snap matches or whatever you can link into this. The key there is the but is disproportionate. Decisive economic income groups and disparity populations and in the southeast particularly I mean specifically Hispanic groups and African-Americans and so one of the problems there is that there's a history of distrust with the research and with the United States we're pretty familiar with things like to state and other incidents where we've mass the trust of the populations that we're trying to set up and so the kid here was how do we create a framework that would engage the population that we wanted to recreate them that we could that they would trust what we were doing so one of the key things about the whole project is that the entire up good and kind infrastructure the building for monitoring cardiovascular behavior and symptoms is open source and it's also covered about by the populations abroad trying to as well so we actually train arguments from the communities to become part of the project and to infer that goes it's a very insular infrastructure and this is a man I don't understand that it's disingenuous to say that somebody has to my princess cared really community and then teach them how to read it it's like me publishing my scientific research in Chinese or Arabic and you're not a Chinese and Arabic speaker and I say let my results you can see for yourself that it makes sense so the point is that we want to engage the communities and kind of develop the head of infrastructure with them so this is.

[00:05:21]
Trying to do it and we've done this through a series of rock shops. The thing is not a phone it provide a human centered approach for integrating electronic health record data with data when the rats are at risk of incorporating data through the small. Protocols developed methods and a framework for evaluating accuracy of the day that's the key elements of this refined that most of the data going Selectric medical records as writers and the tree data that you capture in the world is nationally noisy most of the stuff that you are if you've got an apple ratch or whatever device on your own and you work around with that music to measure your high right you're fine most of the time it's reporting an inaccurate hot right if you have a competitor or a real cardiac device it's mostly just guessing what you have right is based upon your physical activity so rather things are going to build into this is a framework for understanding and knowing when to trust the bedroom when to discard it.

[00:06:20]
Creating a spot of base a cloud by simple structure to allow the data to be stored in a clown on the rise in the cloud and collected in a state of the right if you like this is through Braniff and after that. Run I P address at Brown University to not be out of scale this is the key is branding a step in the infrastructure of it's abstracted a right from any better technology and then most importantly provide an education and community focus average initiative to empower the community in the participants and that's really a key part of this that I think the receive we really push.

[00:07:03]
On a day to day and month to month basis. So. Part of the things that Found out of the design process was. Groups from the communities every every few months and we would have workshops that would last $12.00 or 3 days and we would have them co-design parts of the application with us and suggest things that they would like to see.

[00:07:29]
In the the app itself but they were finding gaging and useful in their everyday lives but other things. We started to develop upon as part of this was to look at. Access this is never really part of the grand right at the start but it's something that found as we went through this project many things the community has identified is really improvements access to good food good quality food and healthy as well as taste and food something that they would enjoy eating.

[00:07:58]
And us are a 3rd of the food so we started to build. A In city after a right of identifying just that the quality of the food rats in your life and neighborhood and we used food as a map from the C.D.C. I mean it's a great day with a crisis in approach to.

[00:08:20]
Information about the quality of various restaurants and grocery store and around Atlanta and both high resolution 3rd man and you'll be at a soup everyone present from here by the way is a part of a series of prices that might represent late afternoon the see me put your hands up guys this and people going over the river a car by him and if you really see and really hear about each of these different parts of the projects and how they're actually coming to fruition these guys I have a nice pair was in there running through it with you and they tell.

[00:08:56]
So revamp this. Ability to tell people how the desert see their location is of the same time appealing. Because this is a factual day to day behavior and so I was just showing some of the serving him. Over the last few weeks the amplitude of my activity has been getting higher and higher and higher and that's not because I've been getting healthier and healthier and healthier and I'm feeling better and better and better it's because the bread has been getting better and better and better and I've been taking the bus last I've been cycling and there's a natural magic lights in the comes from the environment sad how any information about the environment is really important so we put in some real time but of course local pollution affects your day to day activities and that's a part of the rights might affect you.

[00:09:47]
And. Your ability to access the food is a 2nd how to long term cumulative effects of repair of these premises in real time. Recent reports saying back in science frankly integration with. The social network monitoring amount I mean by that is actually looking at the size of the real social network.

[00:10:11]
Looking at how to. How many people are we communicating with in any given day how great and very decent social network and then to start Cruise contracts and claps we're in this in a completely be identified way we never collect any information on who we are talking through.

[00:10:32]
Your eyes for the activities that you're saying about your activities it's a completely previously preserving matter looking that is just the identified versions of the dates and. Created this red cloud application so we have a bunch of web applications like the film. And we push our data up trip.

[00:10:57]
The safe remembering how we how we collect information on. And from visits from we don't store the actual location of by somebody and sweep that information in real time and then discard the location after we've got that information and we have to keep it relatively OK sions confirm this information we then matters with different individuals behaviors and when the president says afternoon is simply the result Sun City check population half of the population whether sharing the right.

[00:11:33]
Identify the severity of the elements from the Kansas City cut it might help with the questionnaire for the fact the sent their rights we're getting within 5 points on 100 point scale of being out of identify how Rassam the spinning without actually asking the many questions it's just from the actual data from the marble found and issues in this infrastructure some really nice work being done then.

[00:11:58]
Very importantly working a lot of training and community average so every 3 months we have sessions where we broke on training and we work on integrating ideas from the community and building up the infrastructure around the system to serve the community and. The. Research publications that come out of this I'm looking at that from text analysis in a day identified right that struck with the bassy.

[00:12:28]
U.C.L.A.. We rested and some work on the idea with Mission production and I was just telling you about we've been doing some edits were some disparities in AI that have come out of this wreck and. The 3 days of mapping which is is in progress. So finally. Russian opportunities were really keen to share this with the community at large.

[00:12:54]
The whole system is open since. I've been to share the code with other people to use. Really interested in but I have extensions people can make to this particular infrastructure to advantage for other half monitoring. Synonomous and they're looking good. A.P.I. so we can delve into it as well because there's just no end to things you compare from Rabbi P I's if you have location information on your phone and you compare the information.

[00:13:29]
Rather also working on building some programs in data science and health literacy and we're just at the start of this learning from the communities of men we can incorporate in this bad is starting to birth runs and applications now the red hot people explore their own diets so one of the key things for this project is that we want people to see what data were collected from them and then read how to manipulate that data so when I say data science I mean that in a very simple right initially I'm not talking about people putting the end of that data from that's not phoned into tends to flow and then trying to predict whether they're going to have half a day or in 25 years I'm talking about people trying to see that you know there's a strong correlation between.

[00:14:15]
The time that I spent in the gym this week and. The kind of food that I at the next week that it had a positive influence on my behavior and looking at these temporary variations in people's behaviors and had seeking the non health seeking behavior as a master considering the research in food deserts explorations in my socioeconomic income environments and in fact anything that anybody wants to climb on its own.

[00:14:51]
Rope and. Share this with you and. If you have any questions just let me thank you. Thank you Gary. Like I mentioned Gary needs to go his question are you releasing his body is. He giving the 2nd is. To to question sorry for just talking of the microphone which pollutions are going into the 2nd is P.C. few more words of the social network monitoring plan just.

[00:15:30]
In case that's not going on on the broadcast the 1st question was repetition indicates a reply link here so this is the reason and now is the back end for this so it's a stand that. Right I.P.I. that you can connect to and it gives you a job this is just a general.

[00:15:51]
Simulation of a lot of different indexes that a given for this but behind it that rob us of PM 2.5 I was alone and cements when I saw that pulling off you guys remember Corey you wrote this bits to remember what I stepped on off. I think it's just $2.00 and I was on it's recent limited I mean I'd be really nice to see a special resolution isn't fantastic on this of course because it depends on the census you have.

[00:16:22]
Such a project at Georgia Tech but I put the sensors in the back of bicycles and cycling around town and I said I'm going to connect with the president so anybody can connect me with this guy as I would like to build the web A.P.I. that we can play around and get maybe we can get limes.

[00:16:38]
And that and the 2nd question was about say out. Of the social networking. I don't really care people are doing on Facebook and Twitter because that's a really lies to be tired of other people about how bright eyes are but I care about. Your text messaging including how lonely you clearly are.

[00:17:03]
And that's a relative level of social engagement drop over time soon is that during this period its recession network tends to contract this is from an earlier thread to be done in one app writes your social network contract so the number of people you communicate with drops but it also changes so it pivots around and you communicate with a different number of people and the right balance from that is completely different based depending on the type of elements so.

[00:17:34]
Really obvious traits in Baghdad the populations of process interactions and then during the manic phases in these final phases they just they just completely dropped my says there are communication and then it by sound again so that's a really obvious indicator things like you would have a contraction in the social network and then the number of people who are calling and texting and how many outgoing versus incoming call this is a drop and that will contract for a small period of time but I should bounce back Richmond rapidly recovery because mostly you're OK.

[00:18:08]
And you're just burrowed So you're trying to interact with a social network whether it resonates psychiatric problems that doesn't rebound back and you might put it when it's a seeking help from another individual so most of the things we're trying to monitor within the social network it's not what you think of as Madden social record Think of it more as a 990 social network from the phone and you could have done this on your dumb phone and they would have been a very strong indicator and I think it's a better indicator in that way so there's a question I mean.

[00:18:42]
This. Seems to be. Exactly I would love that you know this is. This is a very simplistic. Picture of the explosion so we can get and I would much rather get rich. Out of bounds as Rob because I think a lot of people they need to focus on.

[00:19:10]
In-house expectations or external exposures I'm thinking of noise pollution. You know. See our pollution in the hands. You know everything that can affect your short term and long term kind of ask aioe. So thank you for that offer I'm around to say I'm back on track so let's talk about this after you have it just to give you the if you're just.

[00:19:39]
Interested in the view of this in the book you know the words racial harmony in there not to the conversation and so on but it's by. Its own rules so. That is a. Much more than the we'd all love to go to the I think it's the rest of the biggest so.

[00:20:05]
That that I think has a short term effect on hard labor but of the problems with so I probably. Did ask a signal person than anything else in my life and one thing that. I'm really sure of is that there's no write in December you write positive and negative stress isn't from a cardiovascular point of view except when it's extreme and it's causing a Jackson fraction differences but in terms of high rate variability and even reported isolation changes you don't see any I could be overjoyed at my conversation with you and I could be livid at you and it would have very similar reactions actually they're very strong sympathetic surges which creates very low frequency changes in your heart rate variability a passenger Thetic system starts to get inhibited but they're essentially the same response and so there is an air of it's such it seems like.

[00:21:00]
An easy place to try and sort this response but the difficult part is to some degree writing that in the 2nd part is trying to measure the physiology. Anything. And it principale the balance I said earlier that the data coming from your response is probably largely garbage. In terms of half right and that's just experience of how many millions of pounds you just stick it in months on end of the way into the game so.

[00:21:35]
When you say things like give me something to do with it in the. City. I mean you can you can start to drive down a lot of dates on this you would have to of course be measuring ambient light changes in case that was changing the pupil dilation your son need to know whether your ability to measure the papers was not a function of any on the C. hats for example if you had some diabetic retinopathy or something like that would that affect your ability and so you get confounders from me on this is well I think it's a really interesting N.S.F. research project that some is going to fund very soon but I'm not sure if if it's really very feasible from.

[00:22:22]
A large scale Montreux perspective because you've got a lot of assumptions about how people are holding that small farm whether it's facing that face. I mean I I had one smart phone a lot of different endurance and I use a face amount and sometimes I realize that it's not I'm asking because I haven't got it pointing at my face correctly so I mean you're right about tracking actually read the grid i Pads and i Phones and it's actually not as easy a task is you imagine So it's trying to get proper reactivity is going to be is going to be quite a challenge I think just one.

[00:22:58]
Of them but I think the point right and it's a conversation that we have with doctors about whether we can identify these individual stresses say it's working the bad you know this notion of. Continuously driving the environment and looking at the bio feedback from that they call it.

[00:23:23]
Constantly monitoring social media and I forgot to get the research. Constantly monitoring environment but when you have a very it's not surveillance because bass to many many times I'm at but it's. It's more an anthropological time that I'm blanking on a completely. Just did so much thanks and we look forward to having you this afternoon and I can remember what it was coming next a slight change in the agenda next also from Atlanta Georgia Tech Santiago result talking about smart grids and big data.

[00:24:01]
Fake Thank you. So much for that but they need to speak about they might be to spoke my name is on the card about I am they could be I of this project reach a collaboration that we had a Texas A and M.. It did not include the beaches the P.I. of the project with various other folks Yeah because that was a country as well so let's go big quickly what this project is about on some of the challenges on the bill for $26.00 of these particular spoke so 1st century basically we have a tricity an energy infrastructure go in and moving slowly.

[00:24:44]
Very recently about 2010 or so they we should people about it's not create just over a living a double the electricity infrastructure communications and sensors on compute a sham to make it more modern and flexible reliable RISELEY on so we can achieve a lot of good piece associated We if you think reading level of energy for instance solar energy energy storage electric vehicles we a combined you know power etc happening to them so right 2010 the West I lot of investment from the government implementing the same source in these infrastructure as my meters at home they have been implemented in more than half of they customers are interesting customers in the United States here in Georgia we have a 100 percent penetration prong Georgia Power on site and company.

[00:25:36]
I just say no time out of sensor is in this map that insists on what is called distributed energy resources so they set out to be energy storage etc I know the paddle sophisticated sensor is the sites I think the nice especially measurement units which are very fast sensors that can measure a more than 100 times her psychic per pair perception and can provide information about they real time condition and a state of the power system so it will.

[00:26:10]
Be beta right in marriage was very timely because it read these new sensor capability we have new data for the 5 or there is a money to have more data that we had before much more going to run only a special one day so we can understand a lot of what is going around we have it would be great so for those of you that are not familiar with it we have the desire to get it twisted to greet you so everything is connected the 3 into connection with Ronnie's the rest of the Rockies some other use that use of the Rockies and other money stacks so when I plug my cell phone to that I would that these companies and they're connected to everything else I'm degreed feeling here large bowl power plant there you disconnected in Michigan by pretty miles east on 10 years be very fast and very dynamic they not make system so absurd couple of really bad senses of communication and data provides a lot of capabilities to understand how we can interconnect other energy and address issues such as kind my chanteuse my efficiency etc How come a corporate energy smarts into farms and buildings micro-grid campuses hospitals how can we create with reliable means the efficiency bases government facilities etc Those are the possibilities on this.

[00:27:36]
Kind of 2 types of data on at least one of these kind of more you see or read from some for casting to the mind of recasting V.A. prices etc They really seem to be more data dreaming for them they don't remember but as they are bit by bit on the leaks does it really depends on the physics of the system right and some of the experiences that we have in trying to combine communities.

[00:28:01]
Power system into tricity that these are going to be seem very connection right between when we shop postdocs and computer science in the group Ph D. in power system they really need to work together because it is difficult to understand the underlying physical models all the crucial moments they have of happening in a clear City and I the same time the science behind it so for everything for the challenging for the sun is in process.

[00:28:30]
So we had a great collaboration with the notable center some of those say and I said center is going to be arrested I said be sent to you that piece a.k.a there they have the power system and you need a resource center which basically groups are allowed all the key research areas seeing more than 50 different universities across the across the race.

[00:28:51]
Liberalizing those capabilities that arise in the core and center leader liberalizing the freedom sense that everybody centers etc They have a shiny seam their national weave their bias perhaps seeing in Europe and. In Brazil. A lot up a lot of possibilities in that in that area so this pokies divided by your Strasse associated with if very mobile energy associated with US might meet us on the course to member right which allow us to mystify some of it the meets our.

[00:29:29]
Energy plan is being we say the customer will be shaving these Wait the customer the ability of the customer would respond when you see the day down there a few days is much more complex is much more difficult done than that something of a novel energy say with the storm cellar panels when you go and see the bed type of solar panels or having started as a power electronics stuff a lot of infrastructure a not necessarily what we expected to say thing with PM You were so we're learning a lot from these new data streams.

[00:30:01]
From the from the from the industry but I ask business to synchronize pressure measurement the frass was on an education and our reach. Books are below the rush of conferences I've been prepared to move into the direction of create a nightmare of a sophisticated they are great and maybe have a bill of a number of Rush ups.

[00:30:28]
And there's a big. Conference coming on which he said Besides only the 1st conference AI Tripoli. Of course sponsor last night pizza synchronized measurements that these people use a synchronized measurements are a specialized devices about can measure they a notable times per cycle of 60 hertz multiple times paper 2nd across a wide a yes right across hundreds of miles and they are precisely the synchronizing there to see red in the face or the angle between multiple multiple edness and infrastructure so these will take place in May of the CA A.

[00:31:09]
That is what they did they had to be Air Force and the Iranians they rush up bad we're going to we're putting here in Atlanta we just might need H. on the ladies were shop the 1st worst of all these kind of have a pretty powerful program for the rush up that brings together leaders from your industry and a day I could name weeks on a good morning or good to show the latest.

[00:31:34]
Piece I but it's a big dream which means they have been out of energy is moneyed interests and that to be equals is not building some some connection is still true it was Marcy's So this is a summary or a spoke projects for the challenging endeavor a lot of thinking about over the architecture a rod Roddy's What are they struck to go to get the stuff that you destroy as a whole the overall electricity industry in the country stretchy how date out can never ice those objectives and making sure that we have conveyed to us success metrics right resilience economy sustainability is mounted so we can measure those and making sure that the novelties I mean me shut the cover of a clear not being from the build out to freshen the refresh an idea on the creed and then from that functionality to their strategic objectives are a lot of a lot of connected tech sure but all of the greed.

[00:32:33]
Coming from from physical devices of the way up to markets and also if you from Asian fiber I could take her we see a lot of possibilities of the next step right a member of the group is it is a fast and I don't see stem we want to control we have 8 FOS controls from microsecond early sake on the way to a baby operation on all the way to long term decisions brainstems.

[00:32:57]
A discussion indeed as you may be aware of the largest a blackout in de United States history having picked a recurring just a few months ago how can we read a sign. Infrastructure right so it's not probably 3000 people may be on board arse and he's going to be a bark and I leaks the truth about it tomorrow I'm to decision making on the road to prove that he's to get that energy right that will be distributed in energy what is going to be because Chris going to own it it's cetera so I'm very happy to answer any questions that you may have thanks Thank you sun Jagex I think we will do most as an immediate question about we're waiting for the panel discussion turns at Ross and it works if you want to hand your microphone to John.

[00:33:46]
And the next speaker is John Verde who is from the future I was the form again when he gets miked up he's going to talk about collaborative to protect privacy and you state it responsibly. Thank you so much OK everybody's going. So. I got to be brief My name's John Verity I'm here from Washington D.C. as we say I'm from Washington we're here to help.

[00:34:18]
I am super excited to be chatting with you folks today about an initiative that we have started with you know not a long time spoke one of the newer spokes and we're super excited to be a part of this community we're super excited to get feedback on the project and to figure out whether there are points of collaboration so the think tank in Washington D.C. We're optimists about data and we're supporters of the need for privacy protections for consumers and citizens and of the reasons why I excited about this project research collaborative to protect privacy and used it responsibly that we have seen any number of meritorious public programs run aground.

[00:35:01]
Of privacy controversies to get protection critiques and have those concerns under. Research and initiatives that otherwise could help a lot of people we see this at the federal rather we see the state rather we see it in private industry we see it in the academic sector but for this project we're focused on Mr Power these because there are cities and municipal areas across the country some of which are quite large you know you think about the New York City region you think about the same frustration you think about the Atlanta region right.

[00:35:38]
Supervisions in many ways is there as a lot of jazz request was there just some states right but they tend not to be resourced with the kind of data protection and privacy expertise 1st have at the state level and roof repair some folks at cities and communities around the country to try to improve that and that's not a lot so the design of this project is for us to collaborate with municipal leaders to strengthen the.

[00:36:09]
Everybody to collectors and share data on responsible ways one of the things we heard over the past few years of us working in this space is that me this is a reader's did not have a peer to peer network in which they could exchange best practices they could collaborate on overarching projects and cross-cutting initiatives and that they could share practical truths and this is in contrast to other institutions the federal government you have the CIA the council and the CIA the Council brings Chief Information Officers from a crest federal agencies together.

[00:36:47]
As a peer to peer collaboration's to share best practices and assure tools in the private sector you have chief privacy officer said companies who come together in better through IP prove whether through other organisations to share best practices and that for meeting cities the 2nd thing is working with our friends Metro lab networks to try to promote academic partnerships misapply woodies engage and do driven initiatives that often collaborate with academic centers so we're trying to recruit better labs and others to make sure that privacy is baked in and I've seen Hanson technologies are baked in from day one finally we want to make sure that the public understands that when cities and municipalities engage and do drive an initiative is they're doing so in a trustworthy way so.

[00:37:43]
Using the best practices when they're implementing privacy and handsome technologies things in rules that safeguard privacy and maintain the value of data it's important to make sure not just that that's happening but the citizens and communities understand that that's happening because the misunderstanding and the lack of communication can be a real pain point for some of these initiatives particularly at the local level.

[00:38:10]
So. Doing what is stablished the civic duty to privacy leaders network this is the peer to peer network. We've got 13 or 15 cities and municipalities leading folks in the country right now on the ground we have folks from overseas folks in Canada from Ireland who want to be a part of this book spending outside the U.S. to make sure that this can be shared with them that accrue focus is U.S. institutions right and this is everybody from Seattle Washington to Russia.

[00:38:46]
To Columbus Ohio to national labor again to Europe to the sentence right. Folks navigate emerging privacy issues are providing a front for them to share practical guidance. Work with each other to promote fair and transparent data practices. Working with them on a coup or deliverable and the privacy of risk assessment sir privacy risk assessments as some of you are probably aware super important for companies for governments for academic institutions for anybody who correct data that is either individuals for can be linked back into them and they're at the privacy risk assessment is NOT tell our researchers government entities or others that they can't or shouldn't be using personal data rather it's to identify that the potential risk.

[00:39:42]
Is in the sort of information that the steps to get those arrests to have broken is ations balance the potential benefits and the prospective risks in ways that put the trust in the ecosystem and get back from the individuals who state is key to driving these projects. Right so we've done something similar prior to this work.

[00:40:08]
For cities around the program sponsored programs where you have a commitment to transparency and sometimes are under legal requirements to do so that helps. Meet to do that they serve the public and the idea behind that in terms of promoting transparency in government is rather good but there is a pretty good initiatives there service questions when governments collect sensitive personal information about individuals information about the finances information about their health information about the precise location information about their health and safety so we've been working with the question of how do you identify risks identify the benefits craft the balance to get the rest of the benefits into the data space for a while we're taking a lot of those lessons that we've learned and rebirth of those into the privacy risk assessments don't just focus on a computer programs but there are sorts of programs Bermuda's aperitifs correct use and share data.

[00:41:14]
So interesting that outcomes and successes were super excited about their work removing quickly we've got great partner is. Hoping to expand the search through the procedures that work with the process of creating a repository with common tools and expertise for folks to share continue to socialize and publish.

[00:41:35]
Comprehensive risk assessment and we're taking the lessons learned from that to try to transition the privacy risk assessment which applies to not overdo the programs as well and finally requesting workshops with our partners looking forward to collaborating with birds who either have expertise in this group of research going with municipalities through or the folks who are developing implementing privacy and hence in technologies the best practices the can be brought to bear by the members of the network happy to take questions when appropriate during the panel discussion and super excited to work with you folks about how we can promote and drive this sort of the mission of for I mean just with partners around the country around the world thank you John and I Next up we have Franks if you want to take here Mike you hit as you write us a.

[00:42:29]
Fragment of program is the book P.R.I. from the University of Southern Florida talking about a topic that we know relevancy there's a lot of work that has read the inquest of hazards he's going to tout about enhanced 3 D. mapping for Habitat diversity and threat has assessments of question about many areas of the southern US.

[00:42:50]
Concerned and thank you for inviting me over. And I want to thank the N.S.F. for the grant that we have and for the South they to have for getting us to where we are today so I also want to acknowledge my team and I have 3 people from the team 2 post-docs so mad Tyler and Sebastian a student are here with me but we have people on the team from Texas A and M. Corpus Christi and the University of Minnesota so have other people that are at the University of South Florida that are in other departments so Rob we propose to do is to use a commercial satellite data that is no.

[00:43:36]
Variable and a very high spatial resolution that you can get satellite images now that can be less than 2 meter per pixel resolution and not the entire southern coast of the U.S. to develop the 1st ever very high resolution justifications are we going to develop by the minute a special interest in Badlands cost of applications those shows that were just mentioned but in the water.

[00:44:04]
And we want to then drape that those land cover maps on top of a high resolution topographic map that we would also develop over that entire southern coast of the U.S. using light our sort of laser type of term a tree devices so we're also now have the capability of using the commercial satellite data but looking at the story appears to be 3 D. The 3 D. 3 dimensional topography modeling just by looking at the ground from 2 different angles from my satellite so this is the that we're covering.

[00:44:41]
Basically. That the flood plains were interested in looking. Rather to hurricanes to read as are. Prone to flooding and how the people have decided to live and what type of areas what type of vegetation is at risk of further development and and you know how can we help government agencies to to manage the development.

[00:45:08]
So we are going to map the to meet a resolution that us that we have to acquire the data we have to be entered in the license agreement with Digital Globe we're working with it nor are the federal government agencies that have all the light our data for us but they have not pieced it together since all the different formats and.

[00:45:30]
And little snippets of data here and there from different sensors and combine that also into 3 D. and one to progress the not brass are doing very high special resolution maps and I look at a level using a structure for motion using drones so for very small of the red lines or parts of cities we were using that technology.

[00:45:53]
So where we are right now we had a team meeting that we just started in August September of last year so it was pretty new and we discuss projects projects strategy brashly had a test of a drawn collection off of Texas near Corpus Christi and we have now acquired already about 100000 satellite images we have 20000 that bring Manny to collect to the cover of the area of interest we are working with the government to get the light of day that we have to go to bed already for example we have Texas already covered and we're going to start with Texas so we started out in Texas where it goes during dried up projects and we were in favor of doing the move or mapping for it with this project now we are now trying to correct the core to work together to see if we're going to matter in a systematic way and so we started going to start with Texas because they have the light are probably a fee for Texas done and for us it's easy to adapt to that.

[00:46:52]
Right working with the South had to connect with super computer Senator Gregg right now using the University of South Florida supercomputers but we think that we can crank through a lot more data out much more faster working with attack. From when we're also working with Microsoft days or these are examples of the data this is an example of this is Tampa Bay So we mapped on of Tampa Bay a 2 meter resolution so this basically it's a land cover map you can see words with a built environment what's impervious surfaces What's a wetland what's 1st and so on and this is an example of what we did with the drone of Texas that this the breadth of this is probably no more than about 100 meters or so it's a very very very high resolution and over here you have something that may be probably 60 miles across.

[00:47:46]
And this is an example of the topography did have 2 meter 2 meter spatial resolution and border of a about point 2 vertical resolution using LIDAR date on these are applied by your craft So our our goal is to process all of this and we're going to run 2 arrows and so we have to probably reprocess the satellite data out one of these images are independently collecting different dates from different angles from a satellite so things look different and we have to have a maps of look not let not like patchwork so that may have some to be smooth and continuous and accurate so it wouldn't have to reprocess they don't many times and this is very typical with satellite data so having a supercomputer where you can go you can do is is very important for us and then patching of the data together into a seamless maps is also going to take some time.

[00:48:44]
Our goal is to. Distribute the data publicly so the we we are not under license or we cannot distribute the images themselves but our value added product the the land cover maps can we can fully distribute in south that's what people really are interested in and what we would like to do with the hub is.

[00:49:04]
Really for the data so we have we not only have the land cover maps the topography maps at that very high resolution we would like to work with you to find a pic ations and you know if it's specially if it's useful to people. Know some of the things and related to health pollution we may have or be of interest planning for flooding I mean we know that we build the neighborhoods that are easily flooded we've seen that in Houston we've seen that in Louisiana we see it in Florida all the time so right we do it is not completely clear but we do it.

[00:49:36]
So we want to understand biodiversity is we are very interested in carbon mapping and we know that different types of organisms use and process carbon and nutrients differently is not just for carbon so we want to work the issue of biodiversity and using these types of maps. And how to process big data for other biodiversity up locations and if possible not just to the southern coast of the US but the whole world in the sort of eventually we would like to map but tradition everywhere at that type of resolution and do it repeatedly and right now we can't so we look collaboration and understanding how we can do right by making these types of how do we apply these types of data we want to experience and using supercomputers to rescind and have other groups.

[00:50:26]
Use the type of tools that we may develop with the hubs and we also look for experience in mining these types of you know thinking. Thank you for. That. Matter Well I loop. Who is also a stroke Copiague from Spelman College and she's going to talk about her project integrating biological big data research into student training and education.

[00:51:14]
Thank you. All right good morning everyone I am meant to have I am from Spelman College biology department at Spelman we. Just a 4 year college so we only train undergraduates and we're just 10 minutes down the road if there's no traffic of course. I am one of the.

[00:51:40]
Educational focus spokes grant that is integrating Baltica big data research into student training and education and there are further institutions and saved the main P.I. home. At the University of Tennessee Chattanooga and then we have. To. Spelman and then West Virginia University so it's a total of 16 PM I haven't met individually in person all the P.R.'s only through soon meetings or some of them.

[00:52:15]
We are also about you to ha 3 star advisors including grass who is. It's one of the authors undergraduate. Book. It's released by the National Academies of signs some to run who is over to Q.'s cube status for or trying to teach of undergraduate Balochi. Some So I think.

[00:52:47]
Dr Who Tom from Fisk University which is one of the small H B C U's that has launched a science program. And so we are a 1st of all we are a group of faculty who get to do our own research but we also invest a lot of time energy in teaching because we're already working out if we see use are primarily undergraduate institutions so are teaching low tends to be on the heavier side and so through this grant one of the things we attempt to do is to kind of merge the 2 together and integrate our research with our teaching and so that means us teachers scholars we really need to feel confortable with.

[00:53:30]
With the. Science in general but the. Area of expertise and then we usually let that trickle into undergraduate research and eventually into upper level courses and eventually into introductory bargee courses. And so we'll have that dummy expertise and then you know that the quantitative and the competition are thinking built in but we also have been a year and I mentioned you know we have to think of ourselves as poly the rails expertise and the discipline and indeed our science and I chased to have a 3rd bar which would be the education component.

[00:54:13]
And so for example in my south my colleague Dr Lee how I've been participating International Academy stitching fellows are so we've been mentees and eventually became facilitators and these workshops will ban them from developing trust disciplinary teaching modules for example would start module in introductory biology to 1st year students.

[00:54:38]
Would involve modeling and they'll take that let's say and their calculus courses and Mark I'll take that into their introductory. Computer science class and then eventually they really revisit that concept was to have really integrated these different kind of discipline a perspective on the topic. And so the expertise range in Computer Science Biology very broadly anything from ecology I applied to digital imaging ecology called big data.

[00:55:15]
On myself. Plot summary and molecular biologist by training and I have been appropriating some modeling for Mike. And my work one of the big challenges. Undergraduate Education in data science and biology is that the students need to build. Out of knowledge and I would think but they also need to Hobson amount of knowledge in mathematics and computer science and they can only take so many credit hours so how do you really make it happen so.

[00:55:55]
They are spies big data has been oftentimes in the realm of graduate education but so I do graduates need to be somewhat good are savvy so that they are able to move on to graduate programs that are more data intensive. Rap. To the Chinese and the collaborators 40 collaborators are going to be.

[00:56:24]
Accusing again is a plot for him that enables community building around quantity to apology education. A higher state university provides a doll actually health records state on the Jackson lots provided you know make good on and we also have a collaborator here at Georgia Tech who was supposed to give a talk I think earlier but did not make you Iraq hey he's going to be right after we.

[00:56:56]
Did that was he deals with ecology. Beat up. And of course we're connected to a South Park through their priority on education and outreach some of the main activities we propose are vast workshops so we will organize a series of workshops and they're going to have a disciplinary focus more of a technical focus and roles to be for faculty to.

[00:57:30]
Kind of pick up a lot of skill and knowledge and these areas but I also. Developed learning modules with videos so right after delivery of those maybe not you know right away doing the workshops but you know on down the road through all my communities to develop videos that they might use in their courses as well.

[00:57:52]
Another thing really we propose to do is to ask a ha ha Katon So the idea is to provide a lot of expertise over short period of time to faculty who might have really interesting research questions they write top addressed. 2 upcoming road shops I am going to happen to 1st workshops are happening this summer at the University of Chattanooga on the 1st try and be on the don't run wing but company are in the 2nd round on electronic health records and so.

[00:58:27]
To some from a show and would be disseminated through the heart the size hub so these would not be just relevant for faculty who do a lot of tissue but in faculty who actually does he she and then down the road will rush out of the hopper town at Spelman for example as well as competitions along makes that's an area of interest for several of us in the biology department we have faculty who are interested in evolutionary biology type of questions dealing with generally speaking we have colleagues who do a lot of.

[00:59:00]
Mix and I myself do also a lot of mix so it organises. To Visit our already mentioned. That is going to be a vehicle for. Creating extending our communities our life and also really kind of the bottom line is that as a group we hope to be able to create a community shared.

[00:59:30]
In addition to providing development opportunities and ultimately we hope that this will translate into more of a critical role of material incorporated in. That release to biological big data. Thank you we didn't do it. As you said the show date from $2.00 to $10.00 will be about a percent or.

[00:59:58]
Yes. And he is that's the ways that P.I. and his team is using the data for environmental sustainability they did a placid AI technology because it's a simple usable useful knowledgeable thank you very much good to be here at I.Q. this ready shot but I need but eye contact I drew a smile Spitz said that I have a 2 inch we're just a few minutes thank you so the center of a project.

[01:00:29]
We consist of about $25.00 people spread over 12 institutions that include government sponsored institution universities like Dr Detrick as well as industry I.B.M. So it's a big concession. We humans are part of our living ecosystem that consists afterlife of many different friends and the speaker system is constantly changing so we ask questions of the speaker system so here is a Examples are some questions if the radio decided to plan to make a system square faster how many more do it and I expect in my area.

[01:01:09]
This is not just the becoming increasingly important because of an environmental issues including climate to a new group of woman the professional scientists sometimes have a reasonable oncet in discussions and every dicom merits of many of these things are not available but the degree do with professional scientists who are existing discussions citizen scientists and students scientists certainly do not have to do that so the question The becomes and this time after the battle and sensibilities how do we hope that the citizen scientists and of the student scientists male and about ecosystems.

[01:01:48]
So 1st Stephen scientists want them to template scientists like real scientists for citizen scientists citizen scientists to be equally. And that great opportunity to empower them so they can actually build models with you that it's hard to do that so we have developed a research assistant this research assistant uses but that technology as well as big that that I indicated in just a 2nd so a real scientist friend of the cycle if you have an affair pabulum then you create a matter for that particular problem you're right about of your advisor and this is a set of psychographic matter construction really really understood cycle.

[01:02:29]
How do we create such a tool that can support the cycle for citizen scientists with a student scientist so this is a joint collaboration Smythson institution is still writing access to your right. Ear as to its RA just did I miss a biodiversity about 1750000 biological species some more than 60 Jigga bits of data.

[01:02:56]
At the back window here right at the front gate of history spelled modeling and one part of the ad has been to connect them together it's not just a question of failing to do it after everyone. Thinks Smithsonian Dr Dr Jennifer hammock and her colleagues had the right.

[01:03:19]
Making influences on the road the car seat in front mission that mattered to crowds as not necessarily really bring him back fab so it would come in around which number of family to read yours The Guardian feathers that you really need from want me. To show you some examples of this so heinous student has developed a conceptual matter on the left side that about Festivus F. and the Chesapeake Bay It.

[01:03:47]
Usually students humans in Germany are very good at Globen conceptual models from the POV of the perspective we're not equally good at evaluating them. So we build a small compiler translates all that into sense and if a bit of the language of the conceptual model as well as enough of the language of the simulation model will become automatically generated dissimulation free so you as a user as a citizen scientists still have to move the differential equations in step as to process these because they will combine a subset but.

[01:04:20]
That isn't a good example of there's a range made up by step basis this is a roof in the ground Condorcet agricultural and when I still be has been contaminated by better sites somebody is designing the rat race for sort of as a tourist a base would not be contaminated and again you can do dissemination and if you don't like to do the design change then you can go and change now the owners a lot of information but how do you access some of that information so he was I.B.M. screw ups until really you could ask questions like Where do we have a heads of a particular kind eat and I would tune down uses that soon as a back end to pull out your relevant information from the US that fell afoul of it as I could be deaf life and it is important because there's really been a bad it's a kind of a psychological phenomenon we don't even know most citizen scientists most of the scientists will not be removed a basic fact that's really haven't been part of them with a lot of the examinations or the full truth they can actually do their jobs so I mention it to types of users so frustrating scientists this is not a vast system of the resources that should have deprived in the class A drug that it can draw out ecology it's not being used so this is no longer the citizen of that.

[01:05:40]
For citizen scientists move ahead well chaps at Colorado State University as well as at the recent Citizen Science Association meeting and rightly so this is not a difference a for loop a citizen scientists are good not just sitting in a row that. I prefer shops I do about shop so that aspect of it was good of us to have not access we did a data release and I afterwards as a beta release if you were to take this with the documentation anyone can use it up the route to finally come to the most exciting Pad this summer be going to provide you with so at least your best sites are getting about $700000.00 pics a month so we have not confident enough in the event that you live in a team that we can we can predict publicly that as I write a promise that means we can now do many.

[01:06:36]
Public school science and science education because anyone in the world can use it to and whatever way he or she wants to and that raises also have been a bit of challenge just. Since never looked to be found and grew up so the chances are rigid exactly the way we think that robot I was sort of not really slim but it is a bunch of issues homemade people use them how what can we do to support that and how can being could make not just sit in the scientists and students and US but potentially professional scientists are going to support them resist a whole bunch of issues there but you want to appeal to groups or scientists as well as teachers and administrators So for example a driver that we're good early December who ran a workshop for middle school teachers and the Smithsonian to December be going to run a red shop for high school students who go to I do National Museum of Natural History.

[01:07:26]
This one of the South without stopping on me fast and yet of course extremely excited about it because this is an opportunity that comes and once in a while thank you. That was great. So I'd still like to write the other panelists up to join the show at the front of the room someday I'll go Jon Franken until I'm so we can field questions from you have.

[01:07:52]
No right to be late but will still get a chance to ask questions. A couple questions. Might be passed 4th. Just. Look at what's here we need to. Know what was coming but. No questions no love like that was the right quest for the number of the old working fairly large scale groups but the funding was fairly modest to tell you the other typically leverage the other but.

[01:08:34]
Sure broke the we got the stuff for freeze we're in agreement with the private companies and we're also leveraging other innocent. Nor projects certainly and we beg borrow or steal do it any way we can nothing just really work. Stuff doesn't stand. For the build of what. A great.

[01:08:57]
Story questions what are. The roads big empty find the funding we have received from a step from at the front we couldn't see is that fair. So we have assumed our support from I.B.M. And we're also seeing internal funding from drug or tech to pick just most of the data for a bribe and to recommit to a new group of one.

[01:09:25]
So in our chase it's an educational grant and the biggest focus is are on organizing the right shops so we're not me supporting a lot of effort through the kind. Yeah through for us. Very capable I think of executing with the level of funding we have from N.S.F. but that has formed the basis for a couple of proposals that are actually out there active and pending right now being known as N.S.F. energy use so we're looking to expand and multiplying the last.

[01:10:04]
Questions in the room to handle microphone. Once must be good. We're good. And fair but but can you at least say quickly how you mentioned a lot of the ways that people can get involved in maybe some of the folks that have gone on for a while have products so we want to encourage people to also look at what products are out there and can be built on your tools are there any specific tools or things that you would be looking forward to add as a collaboration specifically for your work.

[01:10:48]
Or are there some tools that the community at large can use as well but what would you like to see if you if you could for the next phase what you might be doing for them what current spokes bokes or the ones that just started you know how you how you plan to involve other universities or other types of entity as mine.

[01:11:12]
Yes that's coming to you and I think that in the case of electricity straight to several topics anything that helps the back of creates problem is exposure to he way to think that we having a good get a good transportation as well there because study says the flooding was a problem in Puerto Rico for he says you find out.

[01:11:32]
There no down the electricity so this there is connections we have to work with the internally into spoken up to much about across this post but I think that read lots of opportunities to that they had I.B.M. paid Also I was made I think. You know one of the issues that we have had is.

[01:11:57]
There are storage the amount of thought we're dealing with there is a very large and it's not typical for a normal university computer lab to have that type of resource or the university to offer that to US and Britain with a big data hub to connect to new programs on data storage the like I think is the open data storage network.

[01:12:20]
We have to find ways to protect the data because there are commercial and yet we want to use it and stage in a way that is maybe accessible to other people especially the products and then having access to supercomputers and having us learn how to use those devices this is an important role of the hub.

[01:12:39]
It's been just a few new book stories never just be rolled out if you look at. These millions of compliments. That are so. Significant stories as yours are there's moments of just as you were to make sure that was a good thing. Yeah I guess I won the I'll just reinforce what Renate said about the work being groups of the break up of sat at noon or around going to priority areas but I courage everybody in the audience to seek out the current spoke to seek opportunities for collaboration or working together.

[01:13:21]
But the question here it's more. I want to focus back. I want to focus back on the story just so Colonel E. What did the P.R. its use for actual storage solutions are you turning to I.B.M. for yourselves for their storage solutions or using Microsoft's storage which run so I think the question for the virtual world was were you currently using for storage your other I.B.M. the tools.

[01:13:52]
On storage is a problem and we do need some hope they are yes I.B.M. is hoping but I B M So Bismarck Morandi and out of Singapore waiting access to 2 side of the story supply say I'm going to be great if we as a comedy come which is going to come to some mechanisms that it is state that can be stored and but it could I do this big issue but in Iraq you can put it on top of your route but B. would like this project to contineu beyond bigger vision of the N.S.F. funding and so how we do Dot as a is a question I get going to put a question to get a story but it also slightly separate.

[01:14:28]
So did I answer your question I don't know the answer to question. Whether we were dealing with it as we just by decided right now it's pretty cheap to keep their locally but to have access to a supercomputer and states the data so it's. That's really the issue is how do you put the data in a blow through or it can be processed and be processed and used many many many times without going through a relatively slow network like going to a lab so that's where the piece new concepts girl where you have a supercomputer tied to a very large data storage facility.

[01:15:03]
Which we're going to have access to so that's what we're trying to do is sort of be the interim things bigger storing that locally we just should base this to Minnesota and baton forth with but by mail that's still the fastest and the cheapest. Yeah thank you Beth finally National Science Foundation also from Washington D.C. but I don't ask that question.

[01:15:30]
To say you asked I'm so Frank you had mentioned the data storage need so are the. You know idea you know I get the sense that that the data that you have that is under restriction on is available to you right you create these land mammals but that's not you know I could for I could see the land maps being created to really stop and search as you said they broader air base color.

[01:15:55]
But then that that underlying data being released and no longer within your your your control so I guess what is the longevity of that that you need on that larger proprietary data set for your research. That's a good question were burdened with a photo of your special center so there are major repository there and so there they are funded through N.S.F. to map.

[01:16:21]
The Arctic in the land Mr on the Arctic and Antarctic and as a side project we have this so they are not funded to store the rest of the world and yet we would like to work with an S.F. and that's to have a publicly available at least for science very high special resolution data set it's a very slow to get it out of the commercial databases so it would be nice to have.

[01:16:48]
The the world and or the data holdings from these private. The satellites in a place where you can do science that without having to pay for each image each image is in the 10s of thousands of dollars and we have hundreds of thousands of images to work with just for the Southern US if you can imagine that the cost escalate very quickly and so having about access.

[01:17:12]
For science would be very important and I curse in S.F. and NASA nor U.S.G.S. to get together and talk about how you address this issue so the long term search for the product itself is probably not a big move but the there are layers that we would produce for G.I.S. are believed to be small compared to read actual raw data but if we want to keep the raw data available for a long time so as to make sure that we can do science 102050 years from now that that is an issue and I don't know that private industry has that in mind there but interest in something that you have quickly in that Mr Lee and in long term storage so that is something that has to be worked with industry so many many dimensions to this problem.

[01:17:58]
Was the question yes. Can you give me a one minute to Marshall about your hackathon the audience space when it is the link it's. The 1st one will be off for 2 years from now so we don't have. It. So the idea is that. My colleagues faculty of a species sometimes might have a research question but might feel limited.

[01:18:29]
To expertise and so on so the idea is to invite participants over to the pianist to submit questions and certain I.D.'s and then recruit. Sometimes groups of students they might be graduate students we might be working with. The graduate program. Are sort of power to take for example.

[01:18:56]
Informatics program here is so they might be willing to. Bring their students who might work on some problem and hopefully some of the problems. The big story would probably require a lot of planning ahead right to me to find the problems that and find the problems and pretty.

[01:19:19]
Please join me in thanking you know our speakers and they are not. Going to hand over not US I have no idea what she was. So it's right out here that's what. I just wanted to say one thing is that we switched over introduced everybody else but I did not introduce sin and so Shannon is then a deputy director for the South big data hub and he didn't he modestly did not introduce himself to the time he was a moderator for this time and so now you have.

[01:19:58]
Well it's time to talk during lunch I hope you will collaborate and talk to the people that are here and please join the 1st 2 breakout sessions education and then on data sharing infrastructure so I would these projects that have educational components and resources that are being developed I would love to hear about them see them and also cost promote them so other people can find them after the help and also the data sharing infrastructure pieces that were talked about read it dive deeper into that about how we could actually collaborate.