The ideas in Indie Web Camp are what I was searching for when I proposed a Stateless Distributed Membership back in 2010. A lot of the tools are being developed now. I added a bunch of links from Amber’s talk to the Vimeo page to make it easier to follow. Clearly I have a bit of reading to do.
Jonathan Rajewski, Champlain College in VT
Digital forensics: anything that can read, store or transmit information, that can be presented at trial. Digital is complex. Hollywood glamorizes the forensics process, which is a problem. Ads of what tech can do, new technology always evolving. Public safety officers investigate crimes that involve technology. Problem: backlogs, each state follows different procedures. Court cases illustrate this. It’s a nationwide public safety issue, delays can affect personally in different ways.
What they do: secure communication portals, copy files. This happened in VT. Time is of the essence in forensics: how many devices in a house? what operating systems? Do people know how to deal with this? Complications: all terabyte drives with lots of data.
Forensics Operations center: each of the people have different specialties, training and tools. Another example in Chattanooga. Evidence in house is complex, and school had 4 different iMacs. Using Internet 2 architecture, they connect the local resources with their service.
Demo: using two tools to copy data from devices at target IP, run forensics to determine what pages they’ve been searching, files on the machine. He’s doing triage: not moving whole drive across the network.
Bringing this to reality: SUI Investigator reports to local and virtual forensics centers to develop a solution as soon as possible.
David Lary, University of Texas at Dallas
thRIVE: timely Health indicators using Remove sending and Innovation fro the Vitality of the Environment
Public health report: 3.7 million died due to ambient health pollution, 4.3 million deaths due to household air pollution. This is a Big Data problem of great social relevance: requires bandwidth, wearable sensors and Internet of things, cloud processing, and two-way communiciations for applications to be responsive. If you have a child with asthma, child might have a personal sensor to communicate alerts and other signals.
Think big: bio informatics growing, medical informatics, environmental (weather, etc.) — bring data out of separate silos and combine together. PM2.5 (pollution, an invisible killer) is tiny: less than 2.5 microns (compare with human hair, about 70 microns wide). Wide range of problematic health outcomes from PM2.5.
Cities may have one or two big sensors, 8,000 sites across 50 countries. David wants more data in tighter granularity. What’s ubiquitously available? Multiple satellites, high time precision, Aqua DeepBlue project (satellites) for both an estimate and error bar of data sets. We can then relate that to health outcomes.
Slide of long-term average map of global particulates, 1997 to present. Some areas show dangerous levels (red), e.g., China where lifespan was reduced by 5 years. Small island where lifespan is 27 years due to military bioterrorist weapons testing nearby. Timely alerts between extreme events may be alert to bring inhaler vs visit to the emergency room.
Workflow shows lots of data sources and partners, machine learning, research with VA Hospital, software development for alerts; this is just the beginning.
Next generation will also monitor for fire and drought. Prevention is way less costly than event like great fires. Water and irrigation is another key issue for future. Taking informatics to a whole new level.
Glenn Ricart of US Ignite starts with the limitations of today’s Internet: limited bandwidth, unpredictable response time, designed to drop packets, and that it’s one big flat Internet. Today’s infrastructure is connected via a series of routers between home and the Internet. (Traceroute can show the actual path, in Glenn’s case the path “hops” across the country and back, 32 routers between his home in Utah and a server in California.) This leads to a “weakest link” problem: even if we have a gigabit pipe to our place, we’re slowed down by other points in the path.
Locavore infrastructure: local server that engages services to Internet with much lower latency (delays). “Locavore” defined (Miriam Webster) as one who eats foods grown locally whenever possible.” Not everyone needs gigabit services, but for those who do, this is essential. Services are in headends, also increasingly in incumbent services. (Slide showing what’s commercially available today, a log-log graph showing how things are moving toward higher speed: same amount of time to send more data, or same amount of data in much less time.)
App: On The Same Page for iPhone/iPad — you can see Glenn’s slides as they’re being played.
Upstream: Software-Defined Networks, SDN controllers orchestrate the paths through the network. With this architecture, we’re addressing limited bandwidth, response time and dropped packets. About the one bit flat Internet (where all the performance problelms and security/privacy issues happen): Customized Virtualization for local control, e.g., health info in special protected area, remote surgery with higher priority and lower latency, other channels with characteristics as needed. This will come about in next few years
US Ignite Communities: Tier 4: Gigabit to and from end users. Tier 3: Locavore = gigabit + GENI rack. Tier 2: SDN intercity connections. Tier 1: Customized virtualization. We’ll see lots of demos today and tomorrow of apps at each of these tiers.