Hacker News new | ask | show | jobs
by doodlebugging 3206 days ago
We used Zello when we were down there with a boat. We are just regular people, not part of anyone's navy, just a few guys with a boat that were part of a much larger civilian response to this disaster.

About the Cajun Navy - There were so many people responding to this disaster wanting to help with their time and boats that many times we found ourselves staging to launch but ultimately unable to get in the water because the neighborhood or area we were planning to search only needed 15-20 boats and two or three times that many boats staged for the task. That means dozens of trucks pulling every kind of floating contraption lined up on relatively dry Houston streets and roads and waited in line hoping to be able to take off into the floods and bring people out who needed assistance. When the authorities determined that enough boats had launched for that area, everyone else had to find somewhere else to go. Massive resources were wasted with boats lined up in one spot that could've been actively searching somewhere else.

We ran into a couple of guys from the east coast following Zello Cajun Navy reports who at great expense, drove out to help and had not been able to participate for two days because they were trying to stage in spots where the Cajun Navy was being called up. Once the call went out to the Cajun Navy, boats flooded in from all over and you were nearly guaranteed to have more than you needed. In Orange, we ran into a Texas-based rescue outfit, not the Texas Navy but some group who dressed in military duds. We explained that we were heading into one of the neighborhoods along the Sabine to check a report and asked whether they had already checked it or knew anything about it. They hadn't heard that report or checked that area so we were cleared to go in and the leader of the group asked how we heard about the people needing help. When told that we heard it over Zello, he commented to the effect that the Cajun Navy and Zello was kicking their asses in mustering people and boats.

Fun stuff but the reality of the Cajun Navy effort from our standpoint was that a lot of people wasted a lot of time following their channels and maybe better coordination with other groups would've made a smoother effort.

About the Cajun Navy channels on Zello and Zello itself - Over time the effort evolved into one where each affected area had a dedicated channel to direct boaters to those who most needed assistance. This was not true at the start when the Houston channel had reports from all over the Houston area and eventually Beaumont, Port Arthur, Orange and Vidor too. It was tough at first to determine where the person filing the report was located although there was a website we used to guide us that was actively accumulating requests for rescue and pinning them on an active map. This map evolved to eventually drop many of the requests that were determined to be expired due to rescue already happening or floods receding, etc. We chased reports from that map in the areas that we worked and several times found ourselves following an old report that had not been updated. Obviously, the solution to that is to have boaters radio in and update as they visit addresses. Once the large collection of initial reports from those who had been rescued were tagged or removed the site was much more useful.

A big drawback of Zello in our opinion is that it is vulnerable to manipulation by anyone with the app installed. We kept hearing reports of shots fired at rescue boaters, attempts to steal rescuers boats, rescuers being attacked by residents, etc. and I feel that they were almost all false. We staged at Addicks Dam at the same time reports were actively coming in on Zello Houston channel about shots being fired and boaters were encouraged at the staging area there and later at Bass Pro Shops in Katy to be aware that some rescuers had been targeted. While at Addicks Dam staging area none of us, all Texans and familiar with weapons and gunshots, heard anything remotely like a gunshot. In the days that we were there none of us heard a gunshot at all though the reports over Zello made it sound like you might be taking your life into your hands if you tried to rescue in some of these neighborhoods. A number of rescue boaters were armed and as far as I know none of them had to fire a single shot. I could be wrong though.

Along the same lines Zello was full of reports about imminent dam or levee failures. This got really old as we moved from Houston to Beaumont to Orange and kept hearing the same reports of imminent failure from possibly the same person who kept assuring listeners that the Coast Guard was there with him and informing him that people should leave as failure was imminent. Jeez. Give it a rest. I checked maps around Orange and found that there is no large lake or dam or levee just upstream and we radioed to correct the reports letting people know that they likely applied to Addicks Dam in Houston if they applied to anywhere. Within minutes the same guy came on and announced that the failure of the dam was imminent, the Coast Guard had informed him, etc. A lot of bad information and potentially damaging misinformation was spread.

Anyway. We're home now. Rebuilding will take a while. Help if you can.

4 comments

Massive resources were wasted with boats lined up in one spot that could've been actively searching somewhere else.

I'm on a search & rescue team, and this is usually the chief challenge with "the public" getting involved, there is no established organizational structure. People all mean well, but it is incredibly difficult to utilize a horde of untrained people effectively at the drop of a hat, and it can also make things worse.

Zero criticism of the Cajun Navy et al, sometimes a motley crew is the best you've got and you do the best you can. But if you've ever been turned back when you tried to volunteer for a disaster, this is why.

I wonder if a future extension to ICS/NIMS could be, how to plug in these unconventional resources where professional manpower is insufficient. Develop a 2-minute briefing, a point of contact for unconventional resources to report to, a simplified structure & communication network. The most important concepts are pretty simple, and the first responder doesn't need to know that much about it.

> I wonder if a future extension to ICS/NIMS could be, how to plug in these unconventional resources where professional manpower is insufficient. Develop a 2-minute briefing, a point of contact for unconventional resources to report to, a simplified structure & communication network. The most important concepts are pretty simple, and the first responder doesn't need to know that much about it.

Great idea, but unfortunately when it gets officialized, the lawyers and the bureaucrats get involved and start talking about liability and ways to insert themselves into the situation to profit from it and things quickly gain the complexity of joining the original organization, so it ultimately fails. Maybe I'm too cynical here.

ICS/NIMS trainings are painful to sit through, but in practice it appears pretty agile to me when implemented faithfully.

Instructors will tell you, it's only when you want to be part of the management structure that you really need to know what is going on. Even today, the bottom rung mostly just needs to know where to check in, who they report to, and how to communicate.

I understand your cynicism, and in many places it's applicable. I've noticed however that gov't works developed in the wake of a disaster seems to be pretty solid. Maybe because the project leaders are granted unusual latitude to ignore the lawyers & bureaucrats :)

All things considered, it is in the public interest to allow willing people to spend their time and resources assisting in situations like this. This is possibly the best community-building thing that you can do. Giving everyone regardless of background a stake in surviving something like this can really open up lines of communication that were previously closed by cultural differences or old enmities.

We all came away from this with the feeling that we had done some good for people that really genuinely needed and appreciated outside help. We began actively planning how we could assist the next time something like this happened. We analyzed what we had done and how we did it and identified our own deficiencies and came away resolving to be more prepared.

I think that likely benefits not only us but those who we attempt to assist if there is a next time.

I honestly wasn't trying to knock the Cajun Navy or Zello though it probably reads that way in spots. I'm a consultant in real life and a lot of my work involves identifying logical problems and other deficiencies in my client's operations so that they can focus their improvement efforts where they get the most benefit. Hopefully some of this is useful at some later point.

One thing I forgot to mention that we encountered involves rescuers entering neighborhoods where most residents do not speak English. The 'official' groups we encountered didn't have anyone who could translate English to Spanish or vice versa and thus were unable to understand what problems one potential evacuee needed resolved. Groups should get an introduction to Google Translate or other real-time translation tools so that people who may be in imminent danger can be assisted without having to wait for a translator to arrive.

Houston is a very diverse city culturally with large vibrant populations of people who do not speak or understand English. Once you get into smaller cities you are more likely to find groups where Spanish is the main language. Tools for communication with ALL affected people should be available with all trained, official first responders.

I think your idea about use of unconventional resources is spot on. Most people there with a boat and not part of an official group only wanted to know where they needed to launch or would be allowed to launch and addresses where people were requesting assistance or rescue. I believe a very effective solution to disaster response was demonstrated in Houston and the surrounding area over the last couple of weeks. There are things that could use some polish but overall it is a great model.

One advantage of a trained ham radio operator is that false reports such as you've described are sifted much more thoroughly before being broadcast. Limiting access to the communication channels by requiring licensing is a great way to improve signal to noise ratio.
I agree that trained operators improve the system's ability to handle false input. I'm not sure that requiring licensing for situations like natural disasters is the answer. There will always be people on the ground locally with some of the skills needed to coordinate with dedicated first responders and all of the first-hand knowledge about the status of local infrastructure. Having open, public communication channels for this information to be disseminated probably improves the ability of first responders to direct resources where they are most needed. I think you will always have people ready to inject false or over-hyped reports or who just get emotional and blow things out of proportion. The more people on the ground who can make a report, the easier it will be for a trained first responder to spot and filter reports that clearly are outliers.

Ham operators, I hope, will always be a critical part of the infrastructure. Their ability to relay information when other options are unavailable is very important.

Where there isn't amateur radio available, there's always citizens band. That has a truly terrible (metaphorical) signal-to-noise ratio but it's probably monitored by local police during disasters, not to mention a lot of regular people.
in theory, sure, but unlicensed Ham-capable radios are so cheap nowadays that public repeaters suffer from spam and harassment, just like every other medium.
I have not witnessed much of that, although the couple of cases that I know about were bad for a time.
Why were people lying about rescuers being in danger? Just trolls? I think we would have heard by now if there were pirates out there stealing rescue boats at gunpoint.
Not necessarily lying or trolling. Persons with a low anxiety threshold may post (and re-post) a single report or rumour, not realizing that they thereby create problems rather than assist.
This is probably a part of the answer. Many reports we heard sounded almost like recordings they were so similar. I found myself anticipating some of this dialog just based on previous dialogs. As I noted earlier, when someone tried to inform listeners that the dam failure probably applied to a different geographic area than the one for their channel it was quickly followed by a person repeating that the dam was about to fail, the Coast Guard was right there with him, etc. I decided that someone either had a narrative they wanted to follow or that they didn't realize they were not on the Houston channel any more.
It happened.

http://www.thedailybeast.com/volunteers-helping-harvey-victi...

The National Guard and Coast Guard are armed and did respond to rescuers being attacked. Not sure if they actually fired back in response, or only forced them to leave.

We saw many official groups - most of the swiftwater rescue teams from central Texas, the National Guard, local fire and police departments, organized volunteer groups including the Texas and Cajun Navies, and the Coast Guard. Of those groups the only ones we encountered who weren't obviously armed were the Coast Guard. They may have had their firearms in their vehicles but they weren't on display.

No one that we talked with fired a shot or was fired at. During the several days that we were down there none of us heard a shot even though at times we were supposedly in the neighborhood where boaters had been attacked. We did find some neighborhoods where civilian boats were not allowed - authorities had blocked access except to Coast Guard, Nat'l Guard or other official group boats. Several of these neighborhoods were along the bayous where a lot of expensive homes were built. It seems unlikely that anyone was going to shoot at a rescuer in those areas.

Thanks for pitching in. Can you talk about the "active map"? Who hosted it, who was able to add or remove a report, etc.?
I know of a few people who worked on maps in Houston (and helped some out in minor ways), but AFAIK there wasn't one single map that "won out".

I was the one who created the Cajun Navy's crowd-sourced rescue mapping strategy, though, while helping out with the 2016 Louisiana floods, so I can give a little insight.

At first, I just saw that people were posting requests for rescue in various Facebook groups and becoming quickly buried under requests by others. So I spent a lot of time combing through the posts and adding the details to a spreadsheet, which I then posted on Batchgeo for rescuers to use.

After that, I started working with the people from one of the Facebook groups who had set up a Google Form for submitting rescue requests, and ended up periodically copying requests over from there.

We also had another form for updates regarding a specific request, and were fielding phone calls from people about it as well.

The original map is actually still up here if interested: https://batchgeo.com/map/984218ef8a04f3587f6e723561501e89

Thanks for your efforts. We ended up using a different map that was very similar to the one you linked.
For Houston or Louisiana? The map I linked (and the bulk of my comment) are about what we did in Louisiana last year. I'm not completely in the know regarding Houston.
Sorry. The map you linked is very similar to the map we used for the Houston/Galveston/Orange areas. The people needing rescue were tagged as in your map and useful info about each location was available, in the map we used, on the left sidebar.
I'm pretty sure I stumbled on a map that was linked through an article I read before we left. I believe the article linked to HarveyRelief or HarveyRescue. It may have been a Buzzfeed article like this one:

https://www.buzzfeed.com/maryanngeorgantopoulos/crowdscourin...

The maps are down now as I believe the website went dark once the main flooding danger in Houston was past.

The map appeared to show markers all over the Texas Coast color-coded based on the type of request. The user was able to select a marker and in the left pane a descriptive entry gave information about the address, the number of people requesting rescue including a breakdown of #adults versus #children or elderly. Many posts had telephone contact numbers and each one had an entry from the person requesting assistance granting permission to post their request. Something like an "I Agree" entry after their contact info in the post. I visited the website while it was up and looked at the help request submission part and it was a pretty simple dialog set collecting the needed information along with an "I Agree" or a "Yes" button at the end granting the website the permission to post.

Early problems were predictable. Some of the posts should have been expired because the people had been rescued. Handling the updating of the information required the rescuers to confirm that the people had been rescued so that marker could be cleared. We ended up following several leads that had to be from the previous day. When we arrived at the target address the streets were dry, families were walking pets and kids were riding bikes in the street in one case. A couple of others also involved places where waters had receded and one involved a non-existent address that turned out to be along the beltway between two businesses.

The website did clear many of the older posts and altered the color-coding so it would be clear what you were heading for - medical problem needing assistance, people trapped needing rescue, etc.

It was quite handy. I know there were other maps including one maintained by the Cajun Navy that was similar to the one we used.

There is room for improvement in the maps and in how they coordinate with rescue assets. Timeliness is important and older posts in neighborhoods known to be heavily visited by rescuers should be cross-checked using contact info provided. A lot of traffic on the Cajun Navy channels was people confirming that a rescue needed to happen or that one had already occurred for a particular address.

I believe that the actual requests for assistance on the map I used came from those needing the help since many of them were from someone who was posting for a person needing help who had no access to phone or other communication network. I don't know who had the ability to remove the markers once they were placed but I know that feature by itself needed improvement in tracking rescues that had already occurred.