A couple weeks ago, I published a post extolling the virtues of a nonprofit and open source technology called LDLN. I wanted to highlight the importance of such an endeavor, which is more than most people realize.
After publishing the post, a colleague and long-time emergency manager I greatly respect replied to me, "Whereas I love the fact that you bring new technology to the forefront of disaster management, I often find myself not really understanding what exactly is being discussed. The average non techy emergency manager like myself, who may want to further explore options like LDLN, needs to have an example of its use in the hospital or other environment that is concrete and that can put the technology in prospective."
In reflecting on this, I could have done a better job explaining the problem and how mesh networks such as LDLN play a critical role. This is a complicated but important subject that I want to make sure people understand. So I decided to write another post explaining mesh networks and the value of LDLN.
Mesh networks have been around since the Department of Defense starting playing around with the idea of exchanging data and information in remote and infrastructure-compromised locations. In recent years, mesh networks have been applied to disaster operations to enable the exchange of data and information regardless of Internet access.
However, mesh networks are quite technical to setup and use. A nonprofit and open source technology called LDLN makes this a lot less technical so nearly anyone with some basic tech skills can set up and use a mesh network. Before I dive into how LDLN does this, I want to provide a primer on mesh networks, how they work, and the problems they solve.
The Relationships Betwen Networks, Servers, Routers and the Internet
Let's talk networks, servers, routers and the Internet in a very over-simplified way. Servers are basically supped up computers that can manage the storage of and access to data and information. In some ways, your personal computers access as a server, but when I say server I am talking of machines whose sole purpose is to store and manage access to its data and information. You know that share drive you have access to at work? It is hosted on a server. You know that application that you have access to on the web or only when you are at work? It is hosted on a server. Servers host and store applications with their associated data and information.
In order to access the applications as well as data and information, servers are connected to networks, both wired and wireless. Think of your home network where you can connect your computer, mobile phone, tablet, etc. (also called a "client" in tech terms) via an Ethernet cable or via WiFi. Corporate networks are principally the same, but a bit more complicated in practice. What you need to know is that networks connect you to servers. You rely on this access almost 90% of the time, though you may not realize it. Connecting to your employers wireless network creates an unspoken relationship between your personal computing device and the servers. Outlook is a classic example where the application and the data and information can live on your computer, but all that information is backed up and synced to servers operated by your organization or a third-party vendor.
Now what happens when multiple networks exist or you have to keep an network up across a wide geographic area? It is not so simple for the application with its data and information on your computer to find the relevant server that it needs to sync with. To help direct this digital traffic are routers. The professionals who typically manage this traffic for organizations are called "network engineers." You are an amateur network engineer when you set up your home wireless router, which helps you print to your printer wireless and connect to the Internet. Routers operate in the background to help manage the digital relationship between your computer and servers, printers, the Internet, etc. Routers are especially important when you have many computers and devices on a network that need to exchange data and information. Otherwise, the network would be overloaded and no one would be able to access the servers.
The Internet is like a meta-network that gives you access to the outside world. Many web-based applications live on servers hosted by vendors (or third-party data centers), but are accessible via the Internet because they allow such access. When it comes to Internet access, though, you may have access to your servers via your network, but unless the network is connected to the Internet, you will not be able to access anything external such as web-based applications. For example, you can input patient records into your computer, but you won't be able to access TMZ.com to get the latest dish on Kardashians. So you need to remember that network access and Internet access are related, but separate. You can have network access without Internet access, but not the other way around.
If you don't understand what networks, servers, and routers are and how they work together, the following may be a little harder, but not impossible to follow.
Options for Accessing and Syncing Data and Information
In disasters (and in most of the world), their are generally two ways to exchange data and information no matter what applications you use: 1) a private network, and 2) or the Internet.
Private Network. Before the Internet was a thing, this is where organizations focused their efforts. Organizations set up their own servers, networks and routers at their employment locations to ensure employees have access to and could exchange data and information. All applications along with their data and information remained in complete control of the company and separate from the Internet.
In modern times, a private network plays an important role in data security and control by being able to create a digital wall around data and information (does "firewall" ring a bell?). As you can image, when such an ecoystem is set up with the goal of keeping information in, trying to share data and information across networks becomes extremely challenging. Virtualization and VPNs help mitigate these challenges, but are not perfect and can create some critical and complicated interdependencies. Disaster recovery managers (the IT-focused kind) help plan for and manage these interdependencies so they do not impact operations.
The Internet. The Internet acts much like the networks mentioned above, but in a more public way. Servers are still there and routers help manage the digital traffic in the meta-network called the Internet. The exchange of data and information across the meta-network becomes significantly easier as their are less geographic restrictions. However, using the Internet to exchange data and information creates an extremely critical interdependency. For example, many applications that we have come to love and enjoy on our phones or through our web browser are dependent on Internet access and consume a lot of bandwidth. No Internet means no exchange. Period.
The Problem. These are basically two terrible options for exchanging data and information in a disaster! You can either build applications that work on your private network or build them to work via the Internet. The former limits how far away you can exchange information or across networks, such as hospital-to-hospital or hospital to EOC, and requires the application live on a physical server in your network (e.g., bring a server to the disaster location). The latter creates a critical interdependency on Internet access, which can be a luxury in a disasters.
Mesh Networks in Disaster
Mesh networks allow for the sharing of data and information across wireless networks when no Internet is present. The "mesh" part comes because of the way these networks are typically deployed. A typical deployment model is to "daisy-chain" networks together in such a way that each network shares the data and information with the network it is next to, which then shares it to the next one, and the next one, etc. (think overlapping WiFi signals that link up to each other). And sometimes, if another network has Internet access, you may be able to get Internet access in your network. But the quality of wireless signal drops dramatically the further away you go. Setting up this type of environment is also very technical and difficult in practice!
LDLN and Mesh Networking
LDLN's software and hardware acts as a combined network, router and server. Instead of having to have data and information sync from your computer or mobile device to "the cloud" (servers accessed via the Internet) or internal servers (servers accessed only through your private network), LDLN becomes the best of both worlds. LDLN provides the technology that lets you physically move your phone or computer from one private network to another private network and seamlessly exchange data and information regardless of Internet access.
For example, you have information on your computer that was created while you were in your hospital's network. Now you moved to the municipal EOC that is on a different network, but neither network has Internet access to sync up data and information. With LDLN, as you move to the other network, your computer or mobile devices automatically uploads the data and information on your device to the hardware in that network.
That part is not exactly innovative. What happens next is more innovative:
- If other devices are on the new network, your data and information will automatically be synced to their devices and their data and information will automatically be synced to your devices (regardless of the Internet situation)
- If the new network is connected to the Internet, your data and information will also automatically be synced to the "cloud" for people in other networks to access and their data and information will automatically be synced to your devices.
- If your network is connected to other networks (called daisy-chaining), your data and information will also automatically be synced to those networks and all the devices in those connected networks.
In essence, LDLN has mastered issues that arise when syncing occur asynchronously and distributed across different networks and servers. It will not produce errors when all devices become synchronized with their own as well as each others data. This is huge. Many software solutions don't know what do with conflicting or asynchronous data and information, which causes lots of problems. The software can't reconcile what is the latest information or that it is the same information from different locations such as two receiving hospitals tracking patients.
Gmail handles data conflicts well, but still relies on the Internet for syncing. For example, I might run through my email on my mobile phone while on a plane with no WiFi. I archive some emails, star others, etc. Then I forget and compose an email in my tablet and archive some of the same emails. When I get to the ground, Gmail reconciles what I did on my phone as well as my tablet and doesn't produce any errors. But what if I was in the air with no Internet access and wanted my Gmail on my phone to sync with my Gmail on my tablet? LDLN solves this problem in a disaster environment.
LDLN's Value Proposition
The biggest value for LDLN is to be embedded in various applications, servers and routers. For example, during Hurricane Sandy, this technology could have been integrated with NYU's electronic health records system to share critical patient information stored on their servers with the other receiving hospitals. A person could have physically moved to one of the receiving hospital locations with his or her laptop that had the latest data and information automatically downloaded and synced. That person could have then have electronically shared health records with the receiving hospital. Simultaneously, that person could have kept track of who is at what hospital and have that information be automatically shared back to NYU emergency management personnel.
This is of course a hypothetical example that is over simplified. It merely demonstrates the power of LDLN. Issues such as technical integration, HIPAA and data security would still have to be navigated when setting up this technology. However, I think that can be worked out in the future.
Questions, comments, concerns? I would love any feedback you have on this topic and article.