Where’s my Bus? The Reality of Vehicle Arrival Predictions
Many articles have been written about the advantages of AVL systems in fixed route transit agencies. While AVL does come with a host of benefits, including predicting next bus arrival times, lofty claims about a rider walking out their door 30 seconds before the arrival of their bus simply isn’t the case in most cities. Agencies salivate at the thought of people knowing exactly when their bus is going to arrive, but it’s important to realize that once implemented, AVL delivers predictions and not guaranteed arrival times.
The big question is why? Riders may wonder why, with all the GPS hardware and software, they can’t know exactly where their bus is every second of the day. And, why their PDA says the next bus will be at the stop in two minutes, yet they’re still standing there two minutes after the fact.
To answer this it’s important to understand how arrival predictions are constructed. First and foremost, while a vehicle may know where it is at every given second, it can’t ‘send back’ its location to in-office and online software systems every second. Not only would this incur huge costs (due to the volume of data being constantly sent), but data providers simply don’t allow it (as it would clog up their entire data network). As a result, software packages have to make an educated guess as to where a vehicle is based on the location information it does receive. Typically, vehicle location reports are sent every one to three minutes. If we assume an average vehicle speed of 12 miles/hour, this translates to anywhere from 0.2 to 0.6 miles of distance covered by the vehicle between location reports. That’s 0.2 to 0.6 miles during which the software has to ‘guess’ the vehicle’s location.
In most cases, software applications are able to predict arrival times based on a number of pieces of information:
• Last location received and speed of the vehicle
• Historical traffic and speed information for a given road section
• Distance from the target
• Current traffic information
All of this information does aid in arrival predictions, however, it will never be able to exactly predict an arrival time. Road conditions may vary from calculated values due to accidents, weather, construction, organized events, bad traffic, etc. For example, no software can predict when a vehicle in gridlock is going to start moving again. Sure, the vehicle may only be half a block from its next stop, but it may not move for an extended period of time.
Agencies looking for ‘next stop’ arrival time prediction functionality need to recognize that this information is a prediction—not the absolute truth. This holds true regardless of the software provider chosen.
This is Grace, sales manager from Inelmatic Electronics.
I would like to introduce our company to you. Inelmatic Electronics is a Joint Venture factory of a European and Chinese enterprises, looking for international distributors.
We specialize in customized electronic devices for harsh environments, like vehicles and transportation, including TFT-LCD enhancement and optical bonding.
Besides we design and manufacture high featured, cost efficient OEM sunlight readable monitors and digital signage devices.
More information about us, in our website http://www.inelmatic.com
Our featured products:
http://www.inelmatic.com/web/files/news/XF700_XF1000.html
If you are interested in our services or you want to become our distributor, please don’t hesitate contact me.