ParaPlan Desktop improvements to autoschedule
We added several new features to ParaPlan Desktop this week that make auto-scheduling a large batch of trips more powerful, more accurate and more flexible.
Set MPH for different road types
Based on the length of the trip, we are able to make certain assertions about what types of roads the driver is on between their origin and destination and apply smarter time calculations.
We have five new settings that give our clients a high level of control over their travel time estimations.
-MPH on minor city roads. The default is 17 MPH.
-MPH on major city roads. The default is 30 MPH.
-MPH on highways and limited access roads. The default is 55 MPH.
-Average distance spent on minor city roads. The default is .3 miles, or about four blocks.
-Average distance spent on major city roads. The default is 2 miles.
For example, a typical trip in an urban environment that is ten miles in length is going to start at the client’s house. It will be about four blocks on minor neighborhood roads before reaching a major city street. Once the driver is out of the neighborhood, it will take about two miles to get to the closest limited access highway or interstate. The driver will spend 5.4 miles on the speedier road type before getting off on an exit, driving two miles on a major city road and then another 4 blocks on minor streets to take the client to their destination.
Different agencies have different geographic footprints, and having the ability to fine-tune all these settings means an agency can expect extremely accurate time estimations. For example, because smaller towns lack major roads, a mostly rural agency will set minor road distance to a larger value and set major city roads to a smaller value. A client in a city with well laid out limited access roads like Minneapolis might set the major road distance to .75 or 1 since the driver will be able to spend the majority of a trip on a highway.
Fine-tuned time windows
ParaPlan uses the concept of Time Windows to help our auto-schedule algorithm determine if a trip can be performed on a route. For example, if Betty has a 10:00 AM appointment and it takes her 5 minutes to get off the bus (dwell time), the latest we can drop Betty off is 9:55 AM. To extend that example, if it is a 30 minute trip, then the latest we can pick her up is 9:20 AM (5 min boarding, 30 min travel, 5 min alighting) for the 10:00 AM appointment. It is obviously not very efficient to pick up Betty and drop her off exactly on time, so we have different variables like Riding Constant and Deviation Time to allow a rider to be on the bus a little bit longer in order to accomplish ride sharing. These variables then create the Time Windows.
So back to our 10:00 AM appointment. To extend this example a little further, let’s say we have a 15 minute Deviation Time. We can deviate from the scheduled Pick Up or Drop Off time by that 15 minutes. A client can be picked up a little early so the vehicle can do other work with the client on the bus. So our Pick Up Time Windows would be 9:05 AM - 9:20 AM and the Drop Off Time Windows would be 9:40 AM - 9:55 AM. This gives the bus more flexibility to perform other PU or DO stops as they have a window to hit rather than an exact time.
These variables help, but they don’t allow for late drop offs or prevent too-early pick ups, which are common requests. So we added four new variables to help give further control over Time Windows.
-Amount can be late for Pick Up requests and still be considered on time.
-Amount can be early for Pick Up requests.
-Amount can be late for Drop Off requests (Appointments) and still be considered on time.
-Amount can be early for Drop Off requests (Appointments).
If we were to take Betty’s appointment and apply a 5 minute “Can be late for DO” and a 15 minute “Can be early for DO”. Her Drop Off Time Windows would be 9:40 AM - 10:00 AM. The astute reader will point out that a 10:00 AM drop off isn’t 5 minutes late at all and we would be getting her there right on time. They would be mostly correct, except that it takes Betty 5 minutes to get off the bus, so if we get to the front door at 10:00 AM, it still takes her 5 minutes to get off the bus and into the facility, which would make her Drop Off time 10:05 AM.
Another example would be Ann’s pick-up request of 3:00 PM. We will also assume she has a 5 minute dwell time. If we have a 60 minute PU early time and 15 minute PU late time, the bus could arrive at any time between 1:55 PM and 3:10 PM and still be considered valid.
Max riding time
Extending Time Windows to allow for maximum ride sharing is great, but it is equally important to control the amount of time a specific client can be on a bus. We have a new setting that allows users to set the maximum minutes that a client will be on a bus. Obviously, we can’t apply time changes to trips that are actually longer than the maximum ride time, so if we encounter a riding time greater than allowed by an agency, we will leave it unscheduled and require a dispatcher to manually assign that ride to a bus.