March 2, 2005 By Jessica Jones
Passengers have been inquiring about wireless Internet service aboard the ferries since Jim Long, the WSF's director of information technology, arrived in 2001.
Because of Long's help in implementing the Wireless Over Water (WOW) pilot, commuters on three routes can connect to the Internet from the moment they arrive at the dock, throughout their journey and on the dock at their destination.
What was your role in the WOW implementation?
Washington State Ferries supplied the platform. My role was to work with Mobilisa [a Washington-based wireless company] in arranging for them to get out to the various sites, to review their analysis and to give them our data on our coverage envelopes.
We already have a wireless system for general business, but it's very narrow bandwidth -- nowhere near the bandwidth the customers have. It did not give continuous connectivity and really only gave connectivity, say, a quarter mile from the terminal to the boat. After that, there was no connection. I participated in the WOW evaluation, overall management of the project, managing the funding and reporting to the federal transportation administration. All the technology is done by Mobilisa.
The wireless network is already live on one route. Which other lines will get wireless access, and when?
Right now it's also live on Kingston-Edmonds and on Bainbridge-Seattle. The Bainbridge run is our highest-volume passenger run -- during peak commute runs, we can have 2,500 people aboard the vessel. If 2,500 wanted to connect, that would be a problem. We're thinking anywhere between 100 and 300 could do concurrent work, maybe 500 total.
The analogy I use is Starbucks. You walk into Starbucks -- Starbucks has wireless Internet access, they have 20 customers, and maybe five are using their laptops. That doesn't create a bandwidth problem. But if you've got 2,500 and the same percentage -- or let's make the percentage even less, so 20 percent, or 500 are using it -- that's a tremendous amount of data you're trying to cram down a couple of T1s.
How do you handle 200 or 300 concurrent users?
This is what we're trying to find out. The issue gets to be -- we can get the data off the boat -- it's from its first receipt onshore, getting it out to the Internet service provider. What kind of bandwidth do we really need? When you're cramming things down a T1, which is 1.4 MB per second, how is the ISP going to deal with this kind of volume?
Part of this study is pointing that out. It's not the data on the boat. It's not getting the data off the boat. It's, once it's off the boat and goes through the cloud, if you will, to the initial point of presence, how do we keep that going? When we go out for a formal RFP to the private sector, that will be pointed out.