Promising tools we’ve been looking into include OpenStreetMap (OSM), Ushahidi, and Sahana.

We’re also interested in location-coded use of Twitter (geo-tweeting), any smart phone apps that are field ready, and the use of smart phones in any capacity in the field.

We’d like to ask anyone who is using any of these tools on the ground a few basic, non-technical questions. If you are using any of these in Haiti, or know people who are, please add a comment here or email us at maps2project@gmail.com.

At the time, I knew it had been added to the API, was available for third-party applications, and that there was now a place to enable it in Twitter account settings. It was also listed as a feature in TweetDeck, but I couldn’t find any evidence of it actually working, beyond TweetDeck’s introductory video.

Well, it’s here! At least partially.

Here’s a screenshot of what it looks like in the desktop version of TweetDeck:

Geotagged Tweet in TweetDeck

And the best news is that it’s not that difficult to get started. Here’s what you’ll need to geo-tweet:

• Turn on geotagging in your Twitter account
• Get a Twitter application that can create geo-coded tweets
• Look at your geo-coded Tweets to verify them
• Tell your audience how they can view your tweets on a map

Twitter’s Geotagging Setting

The first step is to turn on “geotagging” in your Twitter acount.

At this time, it seems like this must be done directly on the Twitter website.

Login, then click on the Settings link in the upper right. In the middle of that page, check the box next “Enable geotagging”:

Location Settings on Twitter.com

Then click Save at the bottom. (Don’t forget that part!)

Sending Geo-Tweets

Next, you’ll need a Twitter client application that can create geo-tagged tweets.

So far, there are only a few. I tested two on my iPhone: TweetDeck and Twittelator Pro.

TweetDeck for iPhone is a free app that I’ve been using for several months now. From what I can tell, the geo-features have just been added in the last few weeks. It has two location-related features I really like.

First, you can specify the accuracy of coordinates in metric orders of magnitude, i.e. best, 10m, 100m, 1 km, etc. — an excellent method of obfuscation if there are privacy concerns:

TweetDeck accuracy set to 100 meters

Or:

TweetDeck accuracy set to 1000 meters

TweetDeck for iPhone will also alert you if it thinks the current accuracy of coordinates might be outside the radius of your accuracy settings. After I pressed Send on a tweet with my geo-tagging accuracy set to 100m, it warned me that it had only determined my current location to within 162 meters, and gave me the option of canceling to wait for more accurate coordinates, or sending as is:

TweetDeck Accuracy Confirmation

I once saw this screen when it had only determined my location within about 1200m, so it’s great to know about such discrepancies before sending your tweets.

I also tested Twittelator Pro for the iPhone, which costs US$4.99. (They have quite a few screenshots on their site.) It has some extended audio and video features, so if you’re thinking about sending various types of media from the field via Twitter, it may be worth a look.

Note that the free version of Twittelator does not support geo-tweeting. The options are listed in the settings, but when you try to turn it on, you are directed to buy the Pro version.

Twittelator Pro also has an easy way to embed maps of your current location directly into a tweet using the ‘push-in-crosshairs’ button on the right:

Adding a map link in Twittelator Pro

Adding a map link would be an additional way to share location with audiences that might not be familiar with viewing geo-tweets yet. (See below for more on that.)

Combing through the source fields of a large data set of geo-encoded Tweets, Ed Borasky found a few other possible geo-Tweet clients listed:

UberTwitter, Travel Off The Cuff, Echofon, WordPress, and Adaptive.

I haven’t tested any of these apps directly. If you have experience with them, please share your thoughts in a comment or an email.

Verification

It’s a good idea to confirm that your tweets are actually getting geo-tagged and that the coordinates are reasonably correct before encouraging others to take a look.

Both the apps described above will add an icon to any geo-tagged tweets, and pressing on the icon will display that tweet on a map. For example, in TweetDeck for iPhone:

Mapped Tweet in TweetDeck on iPhone

And Twittelator Pro:

Mapped Tweet in Twittelator Pro on iPhone

In testing both of these apps, I’ve found the accuracy to be pretty good, typically within 20-40 meters of my actual location. I did have one outlier that was coded with coordinates about 600 meters away from my actual location when I sent the tweet, but I was on a moving train at the time.

Here in downtown Portland, we have tall buildings that cause interference, and a dense cell network that helps cell phones determine their location more accurately, so the results of our accuracy tests may not translate to other locations.

We’re eager to hear how accurate these tools are in the field. If you are using location services on an iPhone or any other smart phone/device in Haiti or any other disaster response situation, please let us know.

One caveat: Twittelator Pro will not display any warnings or error messages if the geo-tagging features are turned on within the app, but not enabled for your Twitter account. In other words, you might think you’re sending geo-tagged tweets, but there aren’t any coordinates attached.  TweetDeck does display a warning the first time you try to send a geo-tagged tweet if geo-tagging is not enabled for your Twitter account:

TweetDeck reminder

Tell Your Audience

Since geo-tweeting is so new, there are only a few ways to view geo-tagged tweets so far. I imagine we’ll see a surge in display options and mashups that put tweets on all kinds of maps in the near future.

For now, one of the easiest ways to view tweets on a map is with a desktop program like TweetDeck.

The first image in this post is from TweetDeck, and here’s another one, with the map displayed:

Geo-tagged tweet viewed in TweetDeck Desktop

Why Start Now?

If all of this is so new, why are we encouraging field partners to start using it?

If you are already tweeting, it doesn’t take that much work to start geo-tweeting. Once it’s set up, there really isn’t anything else to do. It just works. (Assuming you have a phone with location services. We’ll explore other options in future posts.)

The minor effort is worth the benefits. What are those benefits?

I can think of a number of them, but this post is already long, so I’ll focus on just two.

There have been several efforts to aggregate and analyze text messages and tweets from Haiti. At the moment, these efforts have had to rely on the location setting in Twitter profiles (often unreliable) or the use of hashtags like #Haiti, which doesn’t necessarily indicate a tweet is really from Haiti.

The ‘noisiness’ of these data sets makes it difficult to analyze them meaningfully.

Ed Borasky took a look at one set of 7500 tweets related to the Haiti relief efforts, and found that only 93 were geo-tagged. As more people start testing and using these tools, the quality of data going into systems like Ushahidi will improve, for this response, and future responses.

As the quality of data input increases, so will the quality of analysis, mapping products and situational awareness tools that can be delivered to the field.

And a more direct benefit: This is a relatively easy way to add location to the story you are telling.

Whether your tweets are directed towards supporters, fans, donors, staff at HQ or the worried mothers of your volunteer teams, adding ‘where’ to your tweets is a great way to keep them all engaged with the important work you’re doing on the ground.

If you are geotagging tweets from Haiti or elsewhere, and have thoughts or questions, please add a comment or send us an email.

• Tradui is a Creole-English dictionary
• Byki has a Haitian French learning app
• Both are free
• I tested them both (on iPhone) in Airplane mode (i.e. no data, wi-fi or voice connections) and they both work fine offline.

Tradui

First up is Tradui, created by Intridea as part of Crisis Camp DC, and is based on the HaitiSurf Creole to English dictionary.

Free download: iPhone or Android

After the splash screen, you can type in a word to translate it from English to Creole, or vice-versa:

Tradui: Searching for food

It doesn’t handle phrases exactly, but you can search for multiple words at the same time:

Tradui: Searching for food

It does seem to be missing a few key words, like clinic:

Tradui: No clinic

Makes me wonder: Is there a list somewhere with the 500 words aid workers are most likely to need that could form the basis for future efforts?

The app also has lists of Creole and English words, broken into folders by first letter:

Tradui: The letter C

There’s a lot of words, which may require a lot of scrolling, and as you can see from the example above, there are some quirks with sort order. The translation screen is much quicker.

I don’t know any Creole, so I can’t make any claims about the accuracy of the translations.

Haitian Creole from Byki

The other iPhone app I tried was Byki’s Haitian Creole edition of its language learning system. This uses ‘lists’ of ‘flash cards’ to
present and then quiz you on different elements of language, similar to a lot of travel language apps.

Byki: List of Lists

Byki: Excuse Me

It could be a useful way to learn before you go, pass the time waiting at airports and on planes, or while waiting at meeting points.

It also has a search feature, which will find appearances of words across cards and lists:

Byki: Water

Though it has some similar gaps to Tradui:

Byki: No clinic

This system is also available for Mac/PC and on the web.

The same caveat applies: I don’t know how accurate the Creole-English translations are.

If you do know Creole, please contact us, and we can pick a few random phrases to test, and report back. Or if you have used these apps, let us know and we’ll update this entry.

Post-Quake Smart Phone Use in Haiti?

At this point, we know of one person who effectively used an iPhone during the Haiti earthquake, but it remains an open question how many responders have them, what the service is like, what it costs, etc.

If you’re an aid worker in the field, do you have an iPhone or Android phone with you? How are you using it? What kinds of apps do you need?

Add a comment or send us an email.

The goal

Ongoing blog posts and direct reports to HumaniNet’s field partners about promising tech responses to the January 12 Haiti earthquake that can help our field partners be more effective on the ground.

Strategies

• Sort through all of the tech activity responding to the Haiti earthquake
• Assess tools/resources/systems according to a standard list of criteria
• Share our results with our field partners

Who we are trying to help (target audience)

HumaniNet Field Partners and other NGOS on the ground (or preparing to deploy soon) who will be involved 10 days to multiple years after the disaster, i.e. well beyond the initial crisis phase.

Inputs

• Needs of NGOs on the ground, and prepping to deploy (determines our research priorities)
• Lessons learned from Maps 2.0 interviews in 2008, simulation support in 2008-2009
• Research leads via blogs, Twitter, emails, personal research

Outputs and Deliverables

• Twitter: raw-ish research, ideas, conversations and promising leads
• Maps 2.0 site: blog posts about individual tools/resources/solutions as we complete them, maybe later posts with updates
• Maps 2.0 site: category round-up pages that point to the individual posts
• Direct reports to field partners and other NGOs that express interest.

Evaluation Criteria

General info we need for all tech we assess:

• How it works (brief description)
• Category (see below)
• Who created it
• Who would use it
• Who benefits from it

HumaniNet’s Big Five Criteria

See this page for background.

1. Provides a capability humanitarian organizations need

2. Potential for widespread use

3. Available to all humanitarian organizations in sufficient quantities

4. Operationally ready for field use

5. Clearly stated costs, valuable to organizations of all sizes

And two new ones:

6. Sustainable

7. Adaptable for Local Use

Security features/concerns

• Authentication? What method?
• Encryption
• – in transit
• – in storage
• – in memory

• Role-based access control?
• Selective sharing between organizations and agencies
• Security audited by a third-party?
• Impact of security protocols on use (i.e. does it get in the way?)

Other Considerations

• Ease-of-use/usability

• How old is it?
• Actively maintained?
• Availability of Support?
• Long-term relevance

• Documentation

• Technical requirements
• Connectivity/bandwidth requirements
• Can it be used offline? With what limitations?
• Reliability
• Interoperability

• Data sources and input formats
• Output formats (KML, CSV, SHP, GeoRSS, PDF, etc.)
• Data portability (Can you get data in and out in standard formats?)

• Data searching and filtering
• Personalized, criteria-based data streams/feeds
• Event alerting via email/SMS

• Costs throughout lifecycle (Startup, operation, retirement and export)
• License fees (or FLOSS aka open source)
• Hardware implementation costs

• Skills needed to implement
• Skills needed for use
• Skills needed for administration and maintenance
• Learning curve for implementation, use and maintenance

Not all criteria will apply to everything we evaluate. Use N/A and Unknown as needed.

Research Categories

• Imagery Sources
• Map sources
• Data sources
• Data collection (includes geo-coding tools/practices?)
• Data analysis
• People-Finding
• Damage assessments
• Operational awareness
• Transportation
• Logistics/Distribution
• Language Tools
• Remote tech support

Additional categories will be added as needed.

If you have any suggestions of additions or modifications, please let me know by emailing maps2project@gmail.com.

Here are Matt Blair’s slides from the event:

• Geo-Usability: Making Effective Maps (13.2 MB, PDF)
• Examples of Maps 2.0 projects in 2009 (10.5 MB, PDF)

Updated: Jim Craner’s slideshow is available on the MapTogether website.

Other followups will be delayed a bit as we do what we can to assist our field partners who are responding to the earthquake in Haiti.

We may do a webcast on these topics, or release narrated versions of the slideshows at some point in the future. If we do, we’ll post about it here.

Thanks to everyone who came out last Tuesday for their questions and enthusiasm!

Clarification: The USGS slides in the Examples PDF are not projects that Maps 2.0 has been working on. I had them in the slides to highlight some of the work  they’re doing such as Did You Feel It? and their Twitter Earthquake Detection Program.

Here are some of Larry’s recent thoughts on charging BGAN satellite terminals:

Solar power will work with the Sabre BGAN but not very well.  A 25W solar blanket will charge it in full sunlight in about two hours with the unit OFF but the unit can run its battery down in an hour of full time use.  You should have access to utility power or else a vehicle battery to support this unit for continuous duty.

…

The Wideye needs DC input 15 – 20 VDC (power adapter and car charger).  The battery is 2.4 amp hour and that shows the unit having ONE hour of transmit time on that battery.  So looks like it uses about 2+ amps of power draw in transmit and about 800 milliamp (3 hours in receive mode) . The solar panel they sell is rated 15-20 volts and 2 amps, so that is 30-40 watts.

Either of the two solar blankets would recharge it when the terminal is OFF, or charge a spare battery but even the larger one may not keep up with the terminal’s power draw in transmit mode, even in full sunlight.

Now the unit may draw a little less than the published specs, and so may last a little longer, but don’t count on it.  For part time use, yes, it will charge it, but for continuous use like this you need more power.  2-3 of the larger ones will run it in transmit mode full time and give the internal battery a charge but will not give enough storage to run more than an hour or so after late afternoon (after peak sun), you will need the second battery if you want to run for hours without external power (solar, utility, or vehicle) to the unit.

This box is pretty power hungry and will need a lot of storage battery over and above the internal one it comes with to run in near 24 hour operation that I expect you will need.

In short solar will work but not as well as you hope for.  The larger one will charge it in full sunlight in about 2 hours with the unit OFF but the unit can run its battery down in an hour of full time use.

I’d recommend that you to have access to utility power or else a vehicle battery to support this unit for continuous duty.

Solar can buy some time during the expected power outages you will have during utility restoration efforts, as the internal battery will go down pretty quickly, but it won’t provide enough power to run and recharge the unit to last overnight or for hours without sunlight.  Running this unit all day will use about 480 watt hours, if you figure 4 hours of full sunlight each day you would need at least 120 watts of solar panels to charge it and about 150 or so amp hours or storage battery to let it run day and night and not discharge the battery too low and damage it.  Sorry for the bad news but I’d rather you be unhappy here before you go than be unhappy in the field with gear and no power for it.

For more information on portable power for disaster response, see HumaniNet’s Power Solutions page. (Some of this material is several years old. We may have an update in the next few days.)

Our goal is to provide our non-profit and NGO partners with specific and actionable information that makes their disaster relief operations more effective and makes their lives easier.

We’ll be researching, aggregating and filtering information for our partners based on our “big five” criteria, our participation in the Cascadia Peril earthquake response exercise and our experience mapping post-quake response activities in Sumatra.

We’ll be posting more in the next few days, so please subscribe by email in the box at right or via our RSS feed to receive the latest.

As a small team from World Vision visits five locations around the world, HumaniNet’s map of the True Spirit of Christmas tour provides yet another way for donors and fans to follow along.

The map will highlight specific reports and photos the team posts to their Facebook page, videos they’ve uploaded to YouTube and items in their online gift catalog.

The comments system built into Facebook allows readers and supporters to share their own experiences and interact with the tour as it is happening. As of this morning, the team is closing in on 10,000 fans on Facebook!

HumaniNet's Map of the True Spirit of Christmas Tour

Matt Blair created the back-end infrastructure part of the map in Portland, Oregon, with design input from HumaniNet’s executive director Gregg Swanson and the World Vision team in Federal Way, Washington. Our super-star Illinois-based volunteer Joe Ludwig will be handling the geo-location and map point entry as the tour proceeds, from Ecuador to Cambodia, then on to Zambia and Ethiopia, before heading back to New York.

We hope you’ll check the map often over the next month or so, and follow along as the tour unfolds!

For information on how to create an easy-to-update Google Map for your project or organization, please email us at info@humaninet.org.

An example of a post

Immediately following the Sumatra earthquake last month, we were able to create a Google Map which presents critical information to responding teams.  You may view this map at http://www.humaninet.org/sumatra-map.html.   Thanks to all who have sent us reports to post, and to Matt Blair and Joe Ludwig for posting the “balloons.”  We are getting very positive feedback on the value of online mapped reports.

As in the Portland exercise earlier this year, we used the Google Spreadsheet mapper.  It works (thanks to Joe) but it has a lot of quirks, as anyone who has used it well knows.

What nonprofit and humanitarian managers now need is a simple system that will enable volunteers and staff to enter data on a Web form, with the information automatically appearing  on a Google Map.  We call this “MapQuick.”  We have checked with our fairly extensive network, including the folks at Google Outreach, and so far have not found anyone that has done this, at least not a system that is truly easy and usable for the average computer user, without middleware.

We envision a system that would be free to nonprofit users.   We have the design and plan for development and have begun a campaign to fund the project.  Stay tuned!

Gregg Swanson

October 21, 2009

Portland (part of Multnomah County). Click for live map.

On April 27, 2009, HumaniNet’s Maps 2.0 team  supported a major disaster response exercise in Multnomah County (Portland), Oregon, by posting essential exercise information on a Web-based online map.  HumaniNet also provided assistance with satellite communications in the multiagency exercise, augmenting landline and radio communications among participants in northwest Oregon.

The mapping and communications events were the first in a series of simulations this year under the exercise name Cascadia Peril, which utilized a scenario of a magnitude 9.0 earthquake followed by a coastal tsunami.

The initial phase, planned and led by the Multnomah County Office of Emergency Management (MCEM),  focused on hospital reporting to the emergency coordination center (ECC) and mapping of events and field reports of affected transportation infrastructure, such as damaged bridges and overpasses.

The HumaniNet Maps 2.0 team worked closely with MCEM in posting time-critical information on Google Earth and Google Maps. The online map was accessible through the exercise at the HumaniNet home page, www.humaninet.org, enabling participants and all interested citizens to observe the exercise activity on a dynamic map. 

The Maps 2.0 team is formed primarily of volunteers who are trained in map-based data entry and who are committed to support humanitarian and emergency response field teams.

The Maps 2.0 planning for the Cascadia Peril exercise built on the experience of the HumaniNet team in support of Exercise Talom in Thailand in November 2008 (see blog post below).