Hyperlocal Mesh Networks

The NY Times has a post up this morning about a neighborhood in Brooklyn called Red Hook where they have built a hyperlocal mesh network to service the entire neighborhood, from housing projects to townhouses.

Red Hook is a cool place. We were there last night to sample Hometown Bar-B-Que‘s massive beef ribs and a bunch of other great stuff. Red Hook is isolated from the rest of Brooklyn by the BQE Expressway and sits right on NY Harbor. It has a collection of different housing situations, from single family homes, to factory lofts, to housing projects. The only public transportation in Red Hook are bus lines into downtown Brooklyn and the occasional Ikea Ferry. It’s a neighborhood all to itself in many ways.

Red Hook was badly flooded in Hurricane Sandy and living there in the weeks after the storm was dicey. The neighborhood has bounced back strongly however and there are construction jobs seemingly on every block. In the wake of Sandy, a local group called the Red Hook Initiative led an effort to build a hyperlocal mesh network throughout Red Hook.

For those that don’t know the difference between a mesh network and a traditional network, the big thing to focus on is that the nodes (think of them like public wireless access points) talk to each other and form a network that operates even if its is not connected to the public Internet. Most mesh networks are connected to the public Internet, but if that connection goes down, the local mesh continues to work. In Red Hook that means that you could make voice calls (over IP) from your housing project to the local hardware store to see if its open. Or you could email a friend who lives in the neighborhood.

If every neighborhood in Brooklyn had a public mesh like Red Hook has, and if they were all meshed with each other, then Brooklyn would have its own local Internet of sorts.

At USV, we think this is an important part of how we (meaning the entire world) get a mobile Internet that is not controlled by the large mobile telcos. We have made one investment in this area (which I don’t think we have announced yet) and we are looking to find other smart ways to invest in this trend.

But the biggest investments that will be made in mesh networking will be made by local groups like Red Hook Initiative. It is not terribly expensive to construct one of these mesh networks and every neighborhood ought to be thinking of doing something like this. If everyone did this, the mobile Internet would look a lot different than it does today.


Comments (Archived):

  1. chris dixon

    Mesh networking could be a great application for Bitcoin. Each node could set its own access rate.

    1. fredwilson

      there are so many great applications of Bitcoin. i wish developers would stop all this alt currency stuff and start building useful stuff on the bitcoin blockchain

      1. William Mougayar

        From your mouth to God’s ears 🙂

        1. Ana Milicevic

          Re: alt currency — could this just be a labeling issue? If we started building apps and calling it something like ‘Net Coin’ for micro payments for digital services perhaps that would drive the immediate applicability over the more long-term currency considerations.@fredwilson @jasonpwright @cdixon

          1. William Mougayar

            The issue is many of these alt-currency ones are attempting to re-invent Bitcoin, or do something marginally better with this or that. That’s a distraction (except for very few cases like Ethereum which is a leap frog type of effort)Point is- Instead of creating a new alt-coin with a new infrastructure/protocol variations for a specific application, it would be better to use Bitcoin proper, and improve what needs to be improved in it.

          2. Ana Milicevic

            Yes, that’s what I meant — build on top of Bitcoin but label this specific micro-payment use case something else so it’s clear of the larger alt currency debate and not bogged down by it.

          3. Tom Clarke

            Ana, you’ve pretty much described Colored Coins, it’s built on Bitcoin but allows the bitcoins to be ‘coloured’ for a specific use… http://coloredcoins.org/

          4. Ana Milicevic

            Thanks Tom! You’re right, that looks spot on.

          5. CJ

            Yeah but then how do you get rich by mining a ton of ‘coins’ before it catches on? We have to see the altcoins fizzle out before development of bitcoin ramps up exponentially otherwise others think they can get rich just by creating a bitcoin clone.

      2. jason wright

        it’s a ‘me too’ gold rush fever.

      3. LE

        stop all this alt currency stuff and start building useful stuff on the bitcoin blockchainA contest would go a long way to getting developers to come up with other applications.

      4. Simone Brunozzi

        Fred, I disagree: the altcoin efforts have the nice effect of stimulating Bitcoin’s own development (see in particular: darkcoin, ethereum, stellar). It’s an ecosystem, after all.

      5. Michael Goldstein

        The vast capital malinvestment in altcoins has diverted resources from the development of some very low-hanging fruit in Bitcoin. For instance, five years later and we are only now seeing the proliferation of multisignature transactions.I’m glad to hear there are investors such as yourself that are devoted to building serious infrastructure for the network, rather than just scamming others in the network for their bitcoins through crowdsales and “IPOs” of economically nonviable software.Great post!

        1. James Rubino

          I agree with you mostly except that Bitcoin was once economically non viable and there needs to be a testing ground where any developer who can contribute something gets a shot at it. The altcurrencies are not state sponsored and epitomizes a freemarket for innovation sponsored by willing participants. Bitcoin can adapt innovations once proven. In the long term Bitcoin has more experience and trust built into it by not adapting every new protocol implementation that comes along. Besides mature bitcoin integration ideas the other best investment is domain experts in the space and you cannot get them if they are pushed to the side either from established Bitcoiners or even people who just do not understand or expect any long term value from bitcoin as a system or currency. In the end the altcoins are community sponsored developers creating an expertise in the space, sponsored by everyday people interested in the Bitcoin protocols. Who else is going to sponsor domain experts??

    2. John Terrence

      True, but on the other hand thought there’s a lot of value in free sharing or even ad based models. It would be a shame if bitcoin destroyed that.

    3. Gregory Magarshak

      Bitcoin is a platform for decentralized zero-trust contracts. It is indeed seeing a bunch of non currency machinery built up around it (like ethereum, maidsafe), but a blockchain is not necessary for most decentralized applications. That’s been around for a long time – distributed hash tables, darknets etc.The thing of course is that the non-blockchain distributed stuff is not zero-trust. You have to trust a server of your choice. But that turns out to be a very useful paradigm. Both from the point of view of the CAP theorem, which limits the efficiency of fully distributed systems anyway, as well as actual usefulness… many applications would be incredibly overkill to implement in a zero trust environment (google Mental Poker for instance). Even a basic chatroom would work better with some central authoritative server per chat. And that stuff’s existed for a long time. That’s how most of the internet looked back before centralized social networks like facebook.But speaking of social… that’s one thing that could use decentralization. So far we’ve had Diaspora as the greatest public attempt. It’s as hard a problem to solve as a decentralized currency. And has just as many (if not more applications). Just as one example out of many… Why is GitHub centralized when git famously decentralized version control? Because of the social stuff. It’s hard to build a distributed client that would support accounts, privacy, and all that other stuff as well as a centralized service. So you get these unfortunate (for the internet) situations where everyone gets pulled into using the same server farm due to network effects, and then engineers scramble to solve the scaling problem and do all sorts of ingenious things that they later post on HighScalability. And a few years later after they have a giant userbase they work to figure out a business model to moneyize the service they provide, which winds up quite successful because of their monopoly position as the platform provider. It’s definitely a lucrative play for venture capital 🙂 but in the end the internet gets centralized around fiefdoms run by google, facebook, apple, etc… I mention all this because as soon as SOCIAL gets decentralized like the blockchain did for money, that kind of phenomenon will begin to get massively disrupted.

      1. Tom Clarke

        There are issues with decentralisation that detract from its potential benefits. Take a look at this story about Diaspora and ISIS for one example of what I’m referring to… http://gawker.com/isis-fina

        1. Gregory Magarshak

          Yes I am aware of this. In fact since we are (have almost finished) building a decentralized platform, I’ve often joked that it would be free for everyone except terrorists. But the truth is of course that it’s more complicated than that. Who should decide who gets to use a tool and who doesn’t? And if you don’t allow them, someone else will build a tool that does. Sometimes it’s better to have them use it rather than go underground — like craigslist with the erotic services section — and allow backdoors (which often exist anyway unintentionally in new software) that can be exploited by the “good guys”. But who is really good or bad in the scheme of things? Douglas Crockford released the canonical JSON library with a modified license saying it shall only be used for good, not evil. And open source distributors had a problem with even this innocuous statement. So Crockford is a bit of a troublemaker… but really, it’s an interesting question.It’s a bit like the question of how do you limit the rate at which someone can brute force a password via a botnet while at the same time not blocking access to the “real” user. If you do one you can’t do the other.I suppose a bettter solution is to have the softwere have multiple ways of detecting when dangerous language is used and voluntarily trying to contact “the authorities” whoever they may be. Rather than allowing wholesale snooping. Maybe with an intermediate step of having some random third party nodes verify that this is indeed dangerous stuff, before starting to distribute it wholesale.I don’t want to say a better solution isn’t possible, so if you hear of anything let me know.

          1. Tom Clarke

            I personally believe the best option is users policing node content (and policing other users, to flag those that are intent on inaccurately flagging content), flagging the content of source nodes in order to allow other nodes to choose not to communicate with them. Wouldn’t stop those directly connected to the node in question but would allow some control over how the content is spread.

    4. @hungoverRyan

      Could this work in cases where the mesh network is not connected to the outside internet?

  2. JimHirshfield

    Interesting. So do users have to manually change the network they’re on… like the way I switch from carrier to WiFi on my mobile phone? Or is all that seamless…some sort of automatic waterfall from cellular to Wi-Fi to mesh, depending upon what’s available?

    1. Drew Meyers

      i’m wondering the same thing

    2. LE

      Answer is here:is that the nodes (think of them like public wireless access points) talk to each other and form a network that operates even if its is not connected to the public Internet. Most mesh networks are connected to the public InternetSo:Your device -> public-accesspoint -> Internet..instead ofYour device -> home-wifi -> Internet

    3. Tom Clarke

      It would be easy enough to make it seamless in the way you describe, using network manager software built for your phone OS. The only drawbacks I can see is that you’d need to keep WiFi switched on all the time in order to detect the mesh network (draining battery life) and it could take a lot of development cycles before you had rapid uninterrupted switching between network sources.

      1. JimHirshfield


  3. Boss Hogg

    Your white space backhaul investment you wrote about a while back (the last SXSW I recall) might complement this — technically, if not as a business. Mesh can get bogged down and white space backhaul links to certain points could relieve congestion.

    1. PhilipSugar

      I was going to post how do you backhaul, but you answered the question. Mesh networks have been around for a long time, but that is a good thing. People forget, that ethernet, wifi, and the internet were around for a long time before they became adopted. .

  4. William Mougayar

    The Decentralization + Unbundling trend continues.Who is the technology provider for these “white boxes” – is it something like http://www.open-mesh.com/?

  5. Dan Kahn

    Also useful in disasters such as hurricanes or earthquakes.

  6. jason wright

    to establish a local mesh is permission from an authority required?

    1. Ana Milicevic

      Which authority would have jurisdiction?

      1. jason wright

        the FCC fascists? 🙂

        1. Ana Milicevic

          :)Distributed network of hyper-local connected communities probably won’t be on their radar for a while. Unless they’re reading this, in which case all bets are off.

          1. jason wright

            such authorities tend to want to extend their footprint of power at every opportunity. it’s the nature of the beast. i hope communities go renegade.

          2. LE

            i hope communities go renegade.You’ll outgrow that. Give it some time.Anyway I thought you weren’t even in the US.

          3. jason wright

            ‘renegade’ for me equates to self sufficiency and empowerment and not being subject to the control of an outside power. a bit like the settlers in North America and their European rulers.you thought right. i’m not. communities exist everywhere. mesh can work well in many countries and cities. the UK gov probably wouldn’t be too keen on mesh because it makes it more difficult to turn off the flow of information if it wanted to do that.i’ve been watching the Ferguson story. i saw Jack Dorsey running around the streets tweeting out what he was seeing. i wonder if the day will come when governments switch off the net in localised areas to control populations and their restive behaviour? could they switch off a mesh? probably have to switch off the electricity supply, so big batteries and solar panels will be needed to counter that.

          4. LE

            Thanks for pointing out the Dorsey stuff, I didn’t know that.I just saw he said this:Dorsey, who more recently launched mobile payment company Square, has also commented on Gov. Jay Nixon’s decision to impose an overnight curfew. Nixon explained it is intended to halt the late-night violence and looting in which some have participated. “I know nothing, but this feels like a missed opportunity to trust our people,” he wrote, directing his post to the governor.I think he is right when he says “I know nothing” that says it all.Let me know if he says anything profound about Isis or the Gaza conflict.

          5. jason wright

            I think St. Louis is his home town.

          6. LE

            Yeah he got his ass out of that place, huh?He thinks he wants to be mayor NYC:http://www.cnn.com/2013/03/…What’s unfortunate is that the way the public thinks (and elects officials) something like that isn’t as far fetched as it should be. Because you know he definitely, um, has the qualifications and experience to run NYC.Kind of reminds me of people who are only interested in something if they can be at the top and aren’t interested in working their way up the ladder. Or maybe setting their sites on something that they can practice at first.I think it’s scary when someone has such an inflated self worth and ego that they think they are capable of running NYC because they were one of the founders of twitter.I’d actually vote for Al Sharpton over Jack Dorsey.

          7. jason wright

            i’m not sure how well a billionaire white boy would be received in Ferguson right now, and he was running around incognito (a full beard and baseball cap) with tweetphone in hand, and refusing all offers of interview from television reporters on the scene. not the typical behaviour of a candidate for public office.i’d vote for Fred, but i don’t have one.

          8. jason wright

            on Isis, the rumour going around is that it’s an MI6 and CIA operation gone wrong. organised and trained to assist the overthrow of Assad in Syria, it then went ‘native’ and is now following an independent agenda in Iraq. as the UK and US had no democratic mandate to send troops to Damascus i can believe that it’s true. why they want Assad out is still not clear to me. why they wanted Saddam out is still not clear to me either.

  7. Ana Milicevic

    This is the way the next currently unconnected billion people will connect to the Internet — using mobile devices of the light smartphone & tablet variety (w/ improved battery life) and not over a centralized backbone. Projects like Village Telco or the Outernet satellite dish are a great start.Deathly curious who you’ve backed in this space. So much potential, especially outside of US markets.

  8. Jim Haughwout

    It will be interesting to see which technology wins out for this infrastructure. While some people are still touting things like Zwave, it appears to be down to Bluetooth LE vs ZigBee

  9. pointsnfigures

    Agree. I don’t think enough people understand the true nature of networks. Professor Ron Burt at the University of Chicago did a lot of research on them in the 1980’s and 1990’s before virtual social networking. I think his research is applicable to any network. I saw a company at TechStars Chicago telnyx.com that looks pretty cool when it comes to disrupting telco’s.

    1. LE

      I don’t think enough people understand the true nature of networks.Who are the “enough people” that don’t understand that should understand?

      1. pointsnfigures

        Mainstream. They understand three degrees of separation, and Malcolm Gladwell’s theory on people-but they don’t really understand the mechanics around them-and the network effects that happen because of them.

        1. LE

          If I understand your point (and I may not) I think that type of thing is less important today rather than more important because of the internet.The example I give is, say, living in a small town vs. NYC (or could be Chicago or Philly). The internet is like a big city with multiple opportunities. Whereas before you had to be careful of what you did and what you said for fear of offending your neighbors (who you might need in a pinch either for help or advice) today, with the internet that is certainly less so. Because there are so many people out there and so much information you are no longer dependent on who is physically close by. Or who you know who knows someone.Before you might need to borrow a drill so you’d have to know your neighbor to do that. Now (hypothetically) that’s not the case. Probably a service now which is “borrow a tool from a stranger”. Don’t even have to check. I know iirc they have things like that. Or they will.Likewise as another example let’s say you are in a certain business in a small town. Let’s say you are an HVAC repairman. If you piss off someone you could loose the whole town. Because everyone talks to everyone else. But living in a large city that’s not the case. For example in NYC (or even Chicago) you could piss people off and as long as you totally didn’t rip them off you could always find new customers because there are so many and people don’t compare notes. Notwithstanding if people check any site which rates service people and remember I did say “not to much”.

    2. Vasudev Ram

      What kind of research did Professor Burt do? Can you summarize it, please?

        1. pointsnfigures

          He looked at social groups within corporate entities. He quantified connections with A LOT of math. He looked at networks, and how to find structural holes within and between networks to enable brokerage.

          1. Vasudev Ram

            Thanks. Interesting.

  10. sigmaalgebra

    > Or you could email a friend who lives in the neighborhood.Well, maybe: So far, usually now there is an e-mail server at an ISP used by the person to receive the e-mail, and from how the Internet works in practice now that server might well not be on the mesh network itself. E.g., e-mail via Gmail would require a connection to a Gmail server, wherever one of those is.Of course, direct person to person (‘peer to peer’) TCP/IP ‘socket’ communications is trivially easy to program and could provide a somewhat clumsy solution to the problem. to be less ‘clumsy’ is part of why we have e-mail servers.But, generally, of course, mesh networks should be able to add some competition to the last mile.Also, of course, wireless itself is an alternative to wires in the ground for the last mile, and some versions should be able to provide competition over the last mile.Yes, as I understand current wireless engineering, there is an issue of bandwidth, but there is a collection of related ideas for cases of spacial ‘beam forming’ to provide more bandwidth in one, small geographic area over the same electronic bandwidth. One outline, just some simple Fourier and matrix playing around, is in my old post http://www.avc.com/a_vc/201

    1. LE

      Of course, direct person to person (‘peer to peer’) TCP/IP ‘socket’ communications is trivially easy to program and could provide a somewhat clumsy solution to the problem. to be less ‘clumsy’ is part of why we have e-mail servers.Oh sure a cat could set that up.But on a serious note your point about the email server is true. There are ways around it but more importantly how much of a need is there for this in other words how often does your internet service stop that people care to even get involved in a mesh network.Not to mention that it also opens up security concerns. Ever been in a hotel with shared wife [1] wifi and been able to bring up files on other people’s computers? I have. Think that everyone has everything battened down on their PC or Mac? They don’t.But, generally, of course, mesh networks should be able to add some competition to the last mile.In theory but you still have to get that connection from somewhere.And the new boss will be the same as the old boss even if they were never the old boss and are a startup. (Always happens when the suits get brought in, that is the corporate types..) And they still have to connect and peer and whatnot. This isn’t 1995 or anything.Also, of course, wireless itself is an alternative to wires in the ground for the last mile, and some versions should be able to provide competition over the last mileI spent more on dinner last night than I did for home internet for the whole month. (Actually I didn’t last night but have). Well maybe when you factor out the cost of TV I did. Anyway. Sort of the same with electricity competition. We’re not talking healthcare here.[1] Don’t say you don’t do this.

  11. BillMcNeely

    would interesting in combination with Nextdoor?

  12. Rob Underwood

    A Community Board 6 (which included Red Hook) committee I serve on, Youth/Human Services/Education, visited RHI and got an overview of this network a couple months ago (later in the meeting, we got to see the Girls Who Code team out there). What’s also interesting is that the network is partially maintained and troubleshooted by local young people in RHI who are getting trained for IT and technology jobs.

    1. fredwilson

      Yeah. There is so much to like about this

  13. diymanik

    Sounds like USV is investing in Open Garden. They’re leading the pack in consumer mesh internet technology with their desktop application and the big success of their Firechat app. Right now they’re the best bet to take mesh networks to the masses.

    1. fredwilson

      Nope. But we have met with them a bunch. I like what they are doing

  14. Carl Rahn Griffith

    Great story. Jane Jacobs would approve.Many of our societal problems stem from legacy hierarchical bullshit. This is disruption at its best.

  15. Tim Panton

    Inspiring, there is a lot to learn from disaster tech. It is worth pointing out that webRTC as implemented in chrome and firefox tries all available routes between peers and has good tech for getting past firewalls. This makes it ideal as an end user client on mesh networks as it will cope with changing topology – plus it is already in > 1bn endpoints so it is ‘pre-distributed’ and pretty secure.

    1. Vasudev Ram

      Have you tried any webRTC video services + clients? Had tried a couple, briefly, earlier, but they had issues, plus I heard the same from a friend. Asking because it seems to be a technology with good potential, if the kinks can be ironed out. Anyone know if G+ use webRTC?

      1. Tim Panton

        Yeah, if you use G+ hangouts in chrome then you are using webRTC (with a couple of hangout specific oddities). The standard hasn’t been signed off yet, but the implementations are now pretty good and still improving – generally a webRTC video call is better than a Skype one for me.

  16. LE

    But the biggest investments that will be made in mesh networking will be made by local groups like Red Hook Initiative. It is not terribly expensive to construct one of these mesh networks and every neighborhood ought to be thinking of doing something like this.Specifically: by local groups like Red Hook InitiativeI don’t want to be Mr. Due Diligence as in “all the guys in HR that I asked didn’t like it so I passed” here [1] but people in the suburbs and in flyover country don’t care about stuff like this. Most places aren’t Portlandia or Brooklyn or Manhattan.[1] Yesterday’s post.

    1. @hungoverRyan

      How can we know they don’t care?Mesh networks may benefit rural/suburban areas even more than urban areas given the greater distance people would have to travel to converse or find each other when internet and telephone service is unavailable.

      1. LE

        How can we know they don’t care?Because it’s human nature and human behavior.Why don’t you do this (hypothetically). Go door to door in some neighborhood in suburban “anywhere” and try to sell households bottled water and tell them it’s a good idea to have it in case the water supply is disrupted or gets dirty.See how many people take you up on it if it’s priced fairly.Next, do the same after there is some scare about the water supply.Not during the scare but perhaps 5 days after the scare. You will sell more water a few days after the scare than you will at any random time when people don’t have a particular fear or need in the back of their mind. (Except of course during the scare). Human nature.To summarize: Out of sight out of mind.

  17. Matt Zagaja

    If I choose to use Google Apps instead of the mesh e-mail provider do I lose that advantage? If I am a start-up and want to have a local node on the mesh network to improve performance is there a cost to this beyond buying the server? If my neighbor streams 4K video from Netflix (or Meshflix) does it overload al the routers?I have to say as nice as wireless is, I just find it is rarely as reliable as wires. When I was in law school all it would take is two classrooms to fill up to start causing slow downs and drop outs. At my current work office we can’t get our wireless to function reliably with 30 employees. All it takes is plugging some Cat5e and I’m humming along without issue.

    1. LE

      If I choose to use Google Apps instead of the mesh e-mail provider do I lose that advantage?What advantage?If I am a start-up and want to have a local node on the mesh network to improve performance is there a cost to this beyond buying the server?Where is the server going to physically sit and where will it draw it’s power from? That will cost somebody something. If my neighbor streams 4K video from Netflix (or Meshflix) does it overload al the routers?Sure. On my work network for example when I do backups from local machines to local machines it impacts performance. Depends on how large the pipes are and what is going when and where.

      1. Matt Zagaja

        The advantage of having my e-mail service available if the Internet goes down.I guess the thing I wonder about is would the collective pay to say have a YouTube or Netflix caching server available because of member demand, but then Fredflix has his own 4D streaming service and has to pay to have his server there because it’s less popular?

        1. LE

          The advantage of having my e-mail service available if the Internet goes down.Well it would still be down if there is no connection to the internet which is the world outside the mesh. Not sure how much utility there is without that.Local cached copies are another story but you can do that now.

          1. Matt Zagaja

            The presumption on my end was that the e-mail service could send e-mails to people/accounts connected to the mesh network without the Internet. Much in the way if the Internet is down you might still be able to e-mail internal people via a corporate exchange server since you’re just connected to it via the LAN.

          2. sprugman

            I don’t think it’s either/or. You would have both a meshmail address bad a gmail one. The meshmail one might be packaged as a chat service.

          3. robertdesideri

            matt, i agree with you’re thinking. years ago, back around 2000, we hacked together a network config to connect meshed and or ad hoc networks and or the internet together for enabling graceful fail over of email transport. for example, should the internet become unusable, for whatever reason, a chained dns system and some open source logic steps in, connecting chain member mail / other servers with all proper authentication and encryption. only modest interest from law firms and a handful from finance, most others thought traditional unencrypted smpt was all any individual / corporate / gov would ever need or want. the sense was no one was going to be sniffing their mail or mess with dns to reroute their packets. today i’d bet many folks believe keeping their email secure and operable during an internet outage is desirable, perhaps even not during an internet outage 🙂 my guess is today there’s an opportunity for a similar plan, hatched by folks more clever than we were -compose some open source logic that fails over gracefully for each member’s mail server (belonging to a dns ‘blockchain’), enabling mail to keep on humming. a dns scheme enabling secure connection anytime, anywhere via any network having access to the chain.

  18. ShanaC

    Red Hook was basically underwater during Sandy.How much of the mesh would actually survive? (edit: mostly because con-ed failed the area, which means the mesh itself would fail without backup power that was waterproof)Or even better, what if the mesh was bisected (not impossible for Red Hook, as it sits near the gowanus canal, on basically sand) (see: power, extensibly of wifi issues)

    1. jason wright

      did Red Hook lose electrical supply due to Sandy?

      1. ShanaC

        Yes. And they’d lose power in another very large storm.

      2. Brook Shepard

        Yes, all of the generators and related Con Ed equipment went up in flames as they were flooded.I’ve lived there, next door to Hometown BBQ, for 7 years. So… It’s a very segregated neighborhood. My block got power within two weeks.The Red Hook Houses (Projects) had no power for ≥ 1 month. National Guard patrolling the streets.

  19. awaldstein

    Fred–is there a grass roots movement of communities building mesh networks to power their local environments? Has an Alice’s Restaurant type of allure around it.Redhook is a cool spot with Redhook Winery doing some innovative work on the natural wine front. My sense is that the city will put in more transportation hubs as there is a ton of interesting things being done on the Brooklyn Navy Yard land.

    1. James Rubino

      I’m no Fred Wilson but having just been reading about this from other sources in the last few days I found some interesting resources.Source on community led meshnets https://wiki.projectmeshnet…OpenBts http://openbts.org/ is a software platform that turns commodity 3g hardware (rural range extenders) into local nodes for networking.Reddithttp://www.reddit.com/r/Dar…

      1. awaldstein

        Thanks James.

        1. James Rubino

          Sure thing. From a developer standpoint OpenBTS looks truly dissruptive in opening cellular networks for local uses and yet somewhat safe for Carriers and hardware providers as they are already integrated in the Business pipeline. Carriers re sell or provide (Femtocells – micro hardware) to rural residents and more capable units to cities.

  20. Vasudev Ram

    OpenGarden – opengarden.com – is doing something like that – mesh networks. They have an interesting take on it. Fred had blogged about it here:http://avc.com/2012/05/open…One of the founders, Stas Shalunov (who was at BitTorrent earlier), participated in the comments on that post. I’m re-reading it now. Post has many comments.

    1. Vasudev Ram

      I had blogged about the progress of OpenGarden a few times – what it is about, funding, tie-ups, traction, etc. Latest post is here (with links to the previous posts):http://jugad2.blogspot.in/2

  21. leigh

    the whole idea of connected local networks with self-healing algorithms kinda blows my mind — I feel like the entire world of ad hoc networks has huge potential — like an on-off web. very cool.

  22. Pete Griffiths

    One application of crytocurrencies has been as a unit of account for bandwidth sharing across mesh networks.

  23. Gregory Magarshak

    I am seriously looking forward to the next wave of decentralization, including mesh networks and energy generation. Not only does it build more resilient communities, but it opens up new forms of governance which make it harder to oppress people (which is the case in other countries).Our company’s building the software for these hyperlocal networks. Like an african village doesnt need drones or baloons and a connection to servers 3000 miles away in order for people to communicate, coordinate, make plans, on smartphones. They can already do that via Wifi access points. There are a lot of applications for a platform like this. Cruise ships. Buildings. Neighborhoods. Organizations of all kinds. Because Facebook/Twitter et al weren’t built for that, and Diaspora leaves a lot to be desired.I spoke with Albert about this. It’s too bad that US regulations don’t legally allow effective wireless mesh networking. The guys at Gotenna ran into that problem. But thankfully we don’t have to deal with the hardware side of it 🙂

  24. JohnEaton2012

    Exciting to think of the possibilities as local communities around the country and world start to build their own DIY mesh networks. An echo of the early Internet days, communities taking back control, building their own mesh localnets.

  25. Jeff Jones

    One of the best BBQ joints in the city. The music space in the back room is great as well. Also check out Sunny’s blue grass jam sessions on Sat nights in Red Hook http://www.sunnysredhook.com/

  26. Jason L. Baptiste

    I think of it like this – if we were to build a new city or colonize another planet, we likely wouldn’t lay fiber underground. We’d use something like mesh networks as they’re cheaper, software upgradeable, and take advantage of everyone being a beacon with devices in their pocket. I undoubtedly believe the Comcast of our generation will be based off of mesh networks. We all have “nodes” sitting in our pockets now.My biggest question has always been around the multi hop problem. There’s been some good research papers on it (http://www.strixsystems.com…, but we’re talking a much larger scale here. This is where a big software advance needs to happen. My one worry is that since it plays into radio frequencies, you’re looking at potential government oversight. Solve this and make it way for thousands of red hooks to spring up and you’re looking at the future.If you’re interested in further reading, there’s a great subreddit here that I follow – http://www.reddit.com/r/dar… .

    1. Jim

      > if we were to build a new city or colonize another planet, we likely wouldn’t lay fiber underground. We’d use something like mesh networksThat is almost certainly not the case. A radio-based network of any sort has a very finite amount of bandwidth available to it, which it must share with all other nearby spectrum users. Depending on the characteristics you want your network topology to have (e.g. distance between nodes, number of nodes in a given area, number of simultaneous connections per node), you typically end up with a rather small chunk of bandwidth available per channel. This is not a trivial problem that you can handwave away, it’s a fundamental limitation of shared-physical-layer systems.You can make tradeoffs when building mesh networks, but there is no free lunch. If you use low power, high density network layouts to maximize channel capacity and prevent interference, then you end up with many hops to cover a set distance. If you use higher power, you get fewer hops, but you also have more clients per node and less capacity per channel.In contrast, when you lay fiber (or coax, or even twisted pair) you have the medium’s entire bandwidth available for a single point-to-point channel. In the case of fiber, this can be a staggering amount of bandwidth because of the frequencies involved. (1MHz of bandwidth at VHF is a lot, in optical terms it is very little.) With the possible exception of space communication where you can simply aim a laser in the right direction and dispense with the fiber, there are no substitutes for that sort of channel capacity.Mesh networks present some very interesting possibilities for building resilient, fault-tolerant, and self-organizing networks, but they are not some sort of magic bullet that will replace traditional directed-graph networks.

  27. Alex Wolf

    Fred, why only buildings? Why not cars like Robin Chase’s idea (the Zipcar founder): https://www.ted.com/talks/r

    1. fredwilson

      oooh, you are on to us!

  28. James Rubino

    If anyone knows Netflix Execs, I would like to pass on the idea that some innovation in mesh nets could be incentivized with one of the catalog boxes the IPs do not want to install one of the meshnets. I’d also like to nominate Austin, Texas as well as Redhook for a testing ground as they both have a fledgling community of mesh netters

  29. laurie kalmanson

    one laptop per child: built on mesh networkshttp://one.laptop.org/about…http://wiki.laptop.org/go/M…there’s also a tablet nowhttp://one.laptop.org/about…

  30. laurie kalmanson

    trolling: war kitteh and denial of service dog; following on concepts of using service dogs to carry wifi hot spots in disasters, using them for the oppositehttp://www.theguardian.com/…

  31. jaginsburg

    If a solution is modular, flexible, scalable, affordable and recyclable, bet on it. It almost doesn’t matter what subject—energy distribution, social networks, personal computers, IKEA furniture—the formula works. I first this little gem from Amory Lovins working on a story for BusinessWeek years ago. And it came up again recently in post about the marvelous Project Frog — I saw CEO Ann Hand speak at the KIN conference (Kellogg Innovation Network). Haven’t been that impressed in a long time… http://trackernewsdots.tumb

  32. kenberger

    (I’m late to this discussion).Grassroots mesh networks is what I focused on from 2000-2003, including a ton of due diligence work for some prominent VC’s. Man, we have some war stories.Here’s a nice piece talking about our efforts doing basically what’s discussed here back in 2002 across the Bay Area:http://www.sfgate.com/busin

    1. fredwilson

      do you think anything material has changed over the past ten years to make the environment more attractive for grassroots mesh networks?

      1. kenberger

        I am more than happy to take a confidential look at your new investment if you want some fresh eyes.I admit I’m battle scarred. Looked at literally hundreds of excellent plans, through 2005. Fon.com may have been the most promising “ground-up” effort using common consumer equipment, and it still hasn’t become very successful. Maybe the 1 improvement you’re asking about is that these days, web- and mobile-based movements and revolutions seem to happen more successfully. Maybe a “new fon” could make it today. The Red Hook project uses non-common equipment– those little boxes, the likes of which have definitely become very cheap. This is a “node” setup, rather than true P2P as with fon.Regulation has *probably* seen a net positive change, if barely, although an issue is this will always vary around the world.You can certainly get pockets of hugely-engaged folks to participate in some locales around the world (I saw it happen 12 years ago). But if you need municipalities involved, or neighboring communities to even talk to each other, that’s the biggest killer. FHP/Tropos (I did early due dili there) pivoted away from that quickly.And then there’s the technology; longer discussion, but other than cost, this area hasn’t really accelerated that much in a long time (frequency hopping was invented by an actor in 1941!), although I might be in for a big surprise. Again, happy to take a look.

        1. fredwilson

          thanks Kenwe will announce our first investment in this space soon and we can continue the conversation then

        2. Silona

          I loved FON even scheduled a meeting several years ago with the CTO of Austin and several city council members during SXSW to discuss a plan and then they didn’t show…

          1. Silona

            Oh yea and I’m late too. But just gave a TEDxBRC talk about a wee bit of this at burningman. (my original talk was too techie ;-))

  33. MC

    It would be awesome if Netflix connected one of their Open Connect servers to the mesh as an alternate distribution method — http://gizmodo.com/this-box

  34. test


  35. paramendra

    This would be amazingly helpful to democracy movements in far away places.

  36. Sascha Meinrath

    FYI — RedHook uses Commotion Mesh: http://www.commotionwireless.net — which is open source software that anyone can download and use for free to set up their own distributed mesh networks (and local applications/services).

  37. LE

    Ask anyone in oklahoma after a tornado disrupts everything. or at the shore when the hurricane hammers everything.a) You need power and after a natural disaster such as you described there is no power it’s kind of one of the first things to go.b) Even if “a” above (gets restored at some point) I don’t agree that these types of things happen often enough that people (and in the same place not state but “where you are” and that people won’t simply lack the umph to set anything up in advance in anticipation. Those are outlier events and they strike intermittently. People are notorious for that type of behavior.So as far as “after a tornado” or “after a flood” of course they will think it’s a good idea. Fred might have had the umph to get a backup generator after Sandy, but within a few weeks or month most people who think about things like that loose that ummph. Now for sure if it happen on a repeat basis that would be a different story. But it typically doesn’t. At least not enough to rally the troops.Same phenomena as the fear of flying after an big airplane crash. Then everybody forgets.