Saturday, March 22, 2014

CAD/AVL (Part 1): History

Three months ago, TriMet's public relations representative Diane Goodwin organized a meeting with me and three men in charge of the development, design and operation of the new CAD/AVL system. They wanted to hold the meeting with me so I could disseminate "accurate" information about the CAD system. This post, and the few that follow, will discuss what I learned in the informational meeting and other sources pertaining to the new CAD system.

Why did this meeting happen? A few months ago, during the roll-out of the new 3100-series buses, I made a post called "The Lost Bus" that told the story of how bus 3159 wasn't accurately reporting its location on the Vehicles data feed, TriMet's Interactive Map, TransitTracker, or even internal systems. Apparently, this post sent waves of mild panic through the IT department as they tried to figure out if there was something really wrong here or if there was a misunderstanding. Turns out, the radio on the bus was not functioning properly and therefore all data was not being transmitted to the system, something that wasn't as major of a problem as it could have been.

Why did it take me so long to write this post? My last post explains a bit about why I have been silent lately. Besides my frustration with all the negativity and my desire to focus my attention on other things, I also started feeling pressure from both sides to post stuff on my blog. Don't misunderstand, I love posting what my reputable sources want me to post, but I don't like pressure to do things and so I just disappeared. So, I am deeply sorry for how long this has taken. I'm not going to let pressure get to me like that again, because life will always be full of that pressure and it does no good to run from it.

Here's how I will go about this. The post you are reading will discuss the background of the CAD/AVL system project and the meeting I was in that provided me with this information. I will follow up this post with more specifics about the system and roll-out in near-future posts. This will help me organize my thoughts and not feel overwhelmed by getting all the information in one post, and it will also allow me to provide information at a rate that is reasonable to the reader. This topic is very complex and is filled with a lot of information, and it deserves a lot of focus because of its importance to TriMet's operations and how greatly it affects riders and operators.

So now I will begin. If you have any questions, comments, or suggested corrections, please feel free to post them in the comments on this page, or you can email them to transitriderpdx2923@gmail.com. Remember, although I know we all have biases and opinions (and I certainly do have my own), my goal here is always that you may know the truth, and where it's lacking, that together we find the truth.

The Meeting
On Tuesday, December 17, 2013, I entered the TriMet building at NE 7th & Holladay to meet Diane Goodwin, TriMet's Public Affairs Manager. Diane has been the one person in the agency who has reached out to me so that I can know what is really going on. She knows that I am skeptical and that I take everything with a grain of salt (from both sides), but she has always shown a level of honesty and integrity that encourages me to trust TriMet a bit more than I would otherwise. In this case, she wanted me to see a demo on the CAD/AVL system, given by the most expert people in the agency on such matters.

The demo was delivered by three men who obviously knew the system inside and out, and who showed the same level of integrity that Diane has shown. A.J. O'Connor, the Manager of IT and Operations Systems, served as the project manager for the CAD/AVL system. He told me about the history of the system: why they decided to implement it, why it is designed the way it is, and details about the actual implementation. Paul Hess is the System Engineer for the CAD/AVL system and provided me with a demo of the system's back office functionalities, showing me what the dispatcher sees and what they can do with the software. John Lutterman is the manager of on-vehicle IT systems technologies and explained how some of the internal systems are designed on board the buses. He also gave me a demo of the "Bus in a Box," a tabletop system that simulates the inner workings of the bus and features an actual touch screen display and a internal stop display. (It was in this part of the demo that I lost my professionalism and started geeking out, smiling incessantly, giggling, and repeatedly declaring, "This is so cool!" Can you blame me? I got to drive a virtual bus!!!)

I also took the opportunity to defend my blogger friends. One of the themes was getting the right information out there so that people are not misinformed. These people were under the impression that my other blogger friends were misinforming the public on purpose because of some hidden agendas. I told them that this wasn't the case, that because the information wasn't out there, we were doing the best we could with what we had. So I greatly appreciated the opportunity to be a voice of honesty and truth to help the right information become available. I was impressed by the integrity I saw in the people at that meeting, and I was reminded that there are still respectable people trying to do the right thing, even in management. It is unfortunate that there are other people who continue to make unwise decisions that make the whole agency look untrustworthy. Remember that there are thousands of people in the agency, and it only takes a few making unwise decisions to take TriMet along a path of disrespect. This is glaringly obvious, as Lane Jensen's second arrest took place literally hours after this demo ended.

CAD/AVL
TriMet obviously needs a system that can track each vehicle in their fleet. There are up to 430 vehicles rolling at one time, and they can't roam freely without some oversight. (Some would argue otherwise, but that's the subject for a future blog post.) So, back in 1995 TriMet implemented a revolutionary system known internally as the Bus Dispatch System (BDS). It was one of the first of its kind in the country, allowing dispatchers to monitor the "real-time" locations of every bus. (This system transmitted locations every two minutes or so, which was remarkable and effective, especially for its time.) It also allowed dispatchers to send messages directly to individual buses or lines about reroutes, police activity, bus bridges, etc. The radio systems on the bus were very old, as that technology hadn't changed in many years. The BDS helped control incoming and outgoing radio communication with the driver, and the two systems were tightly integrated, but they were still two distinct systems made by two different companies at different times.

Between 1990 and 2004, the FCC wrote up two plans to reconfigure the radio frequency allocation (known as "narrow banding" and "rebanding"). TriMet has used two different sets of radio channels for their various vehicles. Buses used frequencies in the 450 MHz frequency, communicating via their 30-year old Motorola radios. The FCC determined that the frequencies were not being efficiently used by public safety and transit agencies, so they instituted a requirement to migrate all radios from using 25 kHz efficiency technology ("widebanded") to 12.5 kHz ("narrowbanded") technology, allowing for more airwave space. All radios made after 1997 could use both technologies, but as TriMet's were so old, they had to upgrade the system completely to meet these requirements. Anybody not in conformance could have their license to operate the radios suspended. (See FCC website, or this FCC PowerPoint presentation.)

TriMet also piggybacked the City of Portland's set of frequencies in the 800 MHz range for their light rail and non-revenue vehicles. This frequency segment was also primarily used by emergency response agencies, and the FCC wanted to make those frequencies stronger and less likely to be affected by non-emergency channels. So the FCC instituted a plan known as "rebanding," where the 800 MHz range was basically rearranged so public safety agencies were all in the same broad range, and private commercial radios were in a different range. This didn't really affect TriMet, it just added to the confusion about all that was going on at the same time. (See FCC website for more on rebanding.)

Remember back in 2009 when we all had to make sure that our televisions either had digital receivers or had that analog-to-digital converter box? Well, the reason the FCC did that was so there could be more room in the analog range for public safety and service radios. TriMet knew they had to vacate the 450 MHz range by the deadline the FCC prescribed, and it made sense to move the buses into the newly vacated 700 MHz range. But TriMet was faced with a choice: should they continue to have two separate radio systems, one for light rail and one for buses, one they essentially rented and the other they owned, or should they have one radio system in one range for all vehicles? As this requirement to vacate the 450 MHz range was inevitably approaching, it made sense to pursue a unified radio system in the 700 MHz range. At the same time, the old BDS system had ceased to be supported by Orbital (the company that made it). TriMet therefore also decided to replace the dispatch system along with the radio system, which was logical since replacing only one system would still leave the need to spend money on the other later (like spending money to replace a motor on a car when the body is rusted out). In summary, TriMet decided to consolidate both the bus (450 MHz) and rail/non-revenue vehicles (800 MHz) radio systems into one system in the 700 MHz range, while also replacing the first-of-its-kind dispatch system that itself wasn't supported anymore either.
TriMet's consolidation of their two radio systems into one.
TriMet then needed to determine whether to contract the systems to two different companies or one who could handle both systems. They did some research, and found out that Seattle had recently implemented new systems and had used multiple contracts. The experience Seattle had with multiple vendors wasn't as great as expected, and TriMet didn't want to face the same problems Seattle had dealt with. So, the decision was made to use one contract for the whole project, thus eliminating any architecture integration and compatibility problems (and, as the saying goes, allowing for "one throat to choke if things go wrong").

The next step was to determine which company would be the main provider. They decided to do field research by traveling to cities who had recently installed new systems and use this experience to help determine the system to use. They travelled to Chicago (which was using a system by Siemens, that has made/will make the Type-2 through Type-5 MAX trains), Tampa (using a system designed by Xerox, which had purchased Orbital, the company that had made the old BDS), and Vancouver, BC (which was using a system designed by INIT, a German company with 30 years of transit expertise). INIT came in with the highest technical score while having the lowest cost, and therefore it was selected to be the vendor. (People have accused TriMet of not including dispatchers in the selection of the vendor. This is untrue, and the people in the meeting validated this statement by telling me that the dispatcher who was taken along on the selection trip was named Ron Dockter.)

INIT then subcontracted the radio architecture design to General Electric, who then subcontracted the actual construction of the radio system to Tait Communications. The decision to have INIT control the overarching project meant that the radio and CAD systems would be integrated as one as opposed to two separate systems. (You can read more about INIT at their website.)

This concludes why TriMet decided to use the system they did. I will post more about what the CAD system does and the actual implementation of the system in the next posts. Special thanks to A.J. O'Conner for his help in providing additional technical assistance in ensuring the accuracy of this post.