Patent Application
20080027772
The invention will now be described in more detail, by way of example only, with reference to the accompanying drawings, in which like numbers refer to like elements, wherein:
The invention consists of an optimization process that unifies three disparate elements of a transit network: vehicles and routes, geographic and demographic regions and funding sources. The data most transit agencies use comes from a variety of internal sources including: schedule databases, automatic passenger counting applications (APC), automatic vehicle location systems (AVL), customer information centers including automated voice systems (IVR) and web-based services, electronic faring ridership surveys, random ride checks, and bus stop databases. External data sources include: census data, map files, National Transit Database information, employment statistics, land use data, school enrolment, ADA clients, and welfare recipients.
This data is relevant to three key areas of transit agency performance: schedule and route adherence and ridership analysis; demographic and location analysis (which portions of the population are or are not being served by transit and what parts of the service area are adequately covered); and demand forecasting (ridership growth and financial planning).
In the prior art, the passenger is the center of the information/data flow as shown in
On a larger scale, the information/data flow continues to operate in the same way. For multiple regions, as shown in
The optimization system of the present invention acts as an information hub as shown in
The optimization system is scalable, as shown in
The optimization system is composed of different modules as shown in
The scheduling module contains all the route and stop information for the network. Locations for bus stops, subway stops, train stops, arrival/departure times and maps of bus and subway routes are all contained within the module. The scheduling module is further capable of analyzing the route and stop data to identify stops with unusually high or low use and suggest modifications to route and stops for optimized passenger capacity and/or ridership.
The dispatching module contains the driver and vehicle availability list, driver assignment and work schedules, and safety and labor (i.e. union contract terms) requirements. The dispatching module takes in the schedule data from the scheduling module and combines it with the dispatching data to create an assignment schedule assigning vehicles to routes and drivers to vehicles. The dispatching module is further capable of handling any other related tasks, including employee payroll records and vehicle maintenance tracking.
The vehicle/driver systems module contains all the data gathered by on-board vehicle equipment for each vehicle and its associated driver. Types of on-board vehicle equipment used include AVL, APC, electronic fare collection, GPS locators, idle monitoring systems, vehicle status monitors and emergency/alarm systems. The module is thus able to provide up-to-date status reports on request, as well as automatically generate alerts and notifications. These alerts can include providing a notice to passengers that a vehicle is running ahead or behind schedule or advising maintenance personnel that a vehicle is incoming for an oil change.
The last module is the passenger interface module. The module contains all of the interfaces used for communicating with passengers. These interfaces typically include a call-center, IVR systems, a website, kiosk or in-vehicle information system for passengers to make route inquires, received optimized route data and report complaints or incidents.
The division into modules presented herein is for ease of presentation and to more accurately reflect the categories of tasks performed by the optimization system. However, in a practical application, there will be many modules handling various specialized tasks that are integrated into the whole system.
The first aspect for optimization is assessment of the number and types of vehicles available on the transit network. This assessment then leads into the determination of available routes. From there, optimal route planning for single passengers, multiple passenger groups and the network as a whole can take place in conjunction with geographic and financial considerations.
Vehicles can be split into two initial types: fixed and variable. Fixed vehicles generally have their routes defined by geography, such as ferries and planes, or by physical requirements (i.e. tracks) such as trains, elevated trains, subways and streetcars. Variable vehicles are generally only limited by the restrictions of available roads and include vehicles such as buses, mini-buses, vans, and taxis. This division of vehicle types is useful for optimization, as changes for variable vehicles (i.e. changing the locations of bus stops) are much more readily accomplished than changes for fixed vehicles (i.e. building a new subway or train station).
While the distinction is made for optimization purposes, the classification as a “variable vehicle” does not preclude the vehicles from operating on a fixed route, like buses stopping at bus stops on a pre-determined route and a “fixed vehicle” may operate on a variable route by omitting stops, like express commuter trains. For most purposes, transit networks use fixed routes for passengers, regardless of vehicle type. However, some transit networks include special networks (herein called “paratransit”) for people with disabilities or other restrictions. These paratransit networks typically use variable routes and lower-capacity variable vehicles, particularly mini-buses and vans.
Most transit operators use a combination of fixed vehicle and variable vehicle services. Furthermore, many passengers may be required to switch vehicles at some point during their journey. For example, a passenger may begin a trip on a commuter train, transfer to a subway, and then take a streetcar or bus to their final destination, with the additional possibility of walking from transfer to transfer, as well as from the bus stop to the final destination point. The optimization system needs to consider the various transfer points and vehicle switches that may be required to travel between any two destinations within the scope of the network.
Another factor is expansion. Most transit operators are constantly evolving, whether by adding new vehicles, building new tracks and stations, or even adding completely new services to the network. The optimization process needs to be able to consider these possibilities when put before it. The impact of a proposed expansion, whether through construction or incorporation of existing adjoining transit networks, should be capable of being assessed by the optimization process.
As an example of the scope of vehicles and routes available in a typical transit operator, the city of Vancouver, Canada has a transit system consisting of a multi-stop single route commuter train, a point-to-point ferry, a multi-stop, multi-route elevated train (SkyTrain™) and a conventional bus system, along with a paratransit network of mini-buses and vans.
Additionally, many public transit operators are required to report on passenger miles, route productivity and performance. Data for such reports come from several sources, including the scheduling database; mobile technologies such as AVL or on-board mobile data terminals (MDTs); APC and electronic faring. The optimization process can be used to correlate this data, preferably using a GIS planning application wherever applicable. At the bus stop level, for example, it is possible to determine the most and least used stops on a route, the number of boardings and alightings at each stop and to compare that with amenities such as benches or shelters. This allows stops to be sited more conveniently and to place the amenities where they are most needed.
Ridership analysis is also essential to understanding the overall performance of a transit system. It allows for the identification of the busiest and least busy trips and those with chronic schedule adherence problems, allowing the transit operator to optimize the overall schedule. Additionally the data can be extracted to determine ridership, passenger miles and other metrics, and automatically generate any required reports. Routes can be spatially analyzed in a number of ways to identify the busiest or least busy times of day, the most appropriate vehicle for the route or time of day and the most and least productive routes. Spatial analysis using maps would also quickly highlight features that affect the productivity of a given route, such as proximity to areas of employment, social services or community attractions.
For fixed route agencies that also provide ADA or similar paratransit services, a route analysis could assist in identifying areas where fixed routes can replace or supplement demand response services. The optimization process should permit transit operators to identify on-time performance short-falls or unproductive routes and then be able to find ways to resolve the problems by adjusting schedules, headways, vehicles or the routes themselves.
Another example of optimization comes from considering smaller transit operators in smaller regions, such as rural areas. By using the optimization process, existing services can be combined to provide greater efficiency and increase transit availability without the need to increase expenditures in terms of additional vehicles and/or drivers. As a result, a level of transit service can be provided which is substantially greater than that currently available in these types of regions.
For example, many communities use school buses to transport children from their home to school and vice-versa. The buses are in operation for certain time periods in the morning and afternoon and occasionally during other times (field trips, sports teams, etc.). By incorporating the school buses into the optimization system, the transit operator is provided with a variety of ways of increasing service through more efficient use of existing resources. One way is to assign school buses to fixed routes that operate in the time periods when the buses are not required for school use (mid-day, night routes). Another way is to add the school buses to the pool of available paratransit vehicles, to cover peaks in demand or routes within the existing school bus route area. Yet another way is to simply add the school buses to the pool of vehicles, making them available to cover emergencies, breakdowns and similar unexpected situations that would otherwise seriously disrupt service.
The next consideration for optimization is geography or, more specifically, the division of geographic regions covered by the transit operators along with the demographic data used to describe each region. All but the smallest of operators will be expected to cover more than one region. Depending on the nature of the operator and the regions, travel from one geographic region to another may be built in or may require substantial adjustments. Handling the passenger transfer from one region to another forms a significant component in the optimization process.
A common scenario is that inter-region transportation is covered by one type of vehicle, such as commuter trains for inter-city transit, and transportation within the region is covered by another type of vehicle, such as buses or subways. With this arrangement, passenger transfer from region to region becomes complicated by the additional need to transfer from one type of vehicle to another.
Outlying regions, particularly rural regions, may have limited or restricted service compared to other regions within the network. These types of limitations must be heavily weighted in optimization adjustments. For example, an outlying region which has only one route available has fewer choices for optimization, however, optimization of the connections to the core areas of the transit network become more significant due to the consequences of missing a connection.
To consider the example provided by Vancouver, twelve municipal regions (Vancouver, Burnaby, New Westminster, Richmond, Delta, Surrey, Langley, Coquitlam, Port Coquitlam, Port Moody, Maple Ridge, Mission) are covered by the transit network, with an additional separate bus system for the regions of North Vancouver and West Vancouver.
In these cities, inter-regional travel is common among transit riders and the optimization process must provide an optimal way for riders to travel between regions, considering all other factors presented herein.
The transit operators must also have a clear understanding of the characteristics of the populations they serve and of the relationship between transit services and the social and economic infrastructure of the community as a whole. Demographic analyses rely on data from the Census Bureau, city, state/province, or county agencies to build an accurate picture of the population's age, income, housing, employment, mobility and many other attributes. The optimization process should be able to work with these data, in their myriad forms, to create statistical and spatial information that can be used to profile existing and potential passengers. Population data from the census or a planning model will be aggregations typically covering multiple-block areas (e.g., census blocks, census tracts, or transportation analysis zones). These data may need to be disaggregated into the areas served and not served by transit. The optimization system can do this and can also be used to match individual address data to areas served and not served by transit. Operators can apply statistical demographic data, such as vehicle ownership or population density, to spatial information such as existing bus routes to create a visual snapshot of who is being served in a given area. The planners can also plot census data on a map, overlay bus routes and create buffer areas around those routes to illustrate the demographic make-up of the service area.
Hand-in-hand with the need to incorporate demographics is the need to incorporate location data into the system. Location data describe where physical elements are, including businesses, other transport services, schools, hospitals, social services, daycare facilities and tourist attractions. By combining information about the population with data about where they travel, operators can build on the schedule and ridership knowledge to use the optimization system to create a truly holistic picture of the transit service as it stands and the changes that might be needed in future.
For example, many agencies in the United States are struggling with transit support for Welfare to Work programs. Only about six percent of welfare recipients own vehicles, and recent job growth has been primarily focused in suburban areas. Using GIS tools, planners are able to identify gaps in transit accessibility and estimate the ability of workers to commute to job locations. A recent study in Boston demonstrated that while 99 percent of welfare recipients lived within one-half mile of transit service, only 43 percent of the jobs in the area enjoyed the same proximity. This type of information is readily available and presentable as part of the optimization process, allowing for quicker response to issues and removing the need for expensive external surveys and consulting.
The optimization system provides tools that allow for detailed exploration by selecting points, stops or polygons to more closely analyze the make-up of a specific part of the service area. These data can then be exported as a report or summary or into another application for manipulation. Operators can also create temporary routes, points and polygons to perform scenario analyses on proposed changes, or they can manipulate the demographic data themselves to assess the impact of population changes on the transit service offering.
In addition to evaluating who is using a transit service, demographic analysis is of enormous value to customer information and marketing departments, who must have an in-depth understanding of their audience characteristics in order to provide appropriate information and services. For example, demographic analysis may indicate that a given route serves a particular language group, and that customer service may need to be offered in that language. For marketing departments that are trying to build ridership, a spatial understanding of vehicle ownership, household income and employment can pinpoint where best to focus marketing resources to promote the transit service. At the executive level, these data are also critical.
The final consideration for optimization is the financial sources which fund the transit operators. A primary source is rider fares. Another source is government funding, typically from taxes, received from different levels of government. Private funding or charitable funding may also be used, particularly for paratransit services.
Rider fares may be fixed or variable, and can be dependent upon several factors, including the type of vehicle and region of travel, as discussed above. Optimization can be used to determine which potential route is the least expensive for travel from point-to-point. Also, optimization can be used to assess vehicles, routes, and/or regions that generate a low number of rider fares and suggest appropriate adjustments. The optimization process may even be used to suggest fare prices, or changes in the fare system and to assess the potential impact of such changes.
Government funding may be local (city/municipal), regional (state/provincial) or federal, depending on the areas serviced by the network and the policies of the government. Generally, funding would be derived from the government's general tax revenues, however, it is also common to have specific tax levies, such as taxes on property, fuel and vehicle insurance which are intended to directly fund transit networks.
Paratransit operators typically receive private or charitable funding and have specific requirements that must be met for a passenger to be eligible. By using the optimization process to assess patterns with the usage of these networks, if may be possible to combine them with conventional transit networks, and the attendant fare charges, to reduce passenger travel time and increase passenger capacity.
The optimization system can also assess where received funds are being used in the transit network and determine if adequate funds are being received for assigned purposes.
Additionally, received funding may be contingent on ridership, and the effects on ridership resulting from the optimization process must be reflected in the financial aspects of the model.
Financial concerns often have a significant overlap with regional concerns. For example, special transit levies, such as fuel taxes or vehicle insurance taxes may only apply to regions serviced by the transit network. Also, regions with limited or restricted transit service may be exempted from these taxes. Additionally, many transit networks use fare surcharges for travel between regions, and the amount of these charges and the definition of regional boundaries are often contentious issues.
By incorporating these financial considerations, the optimization system can be used to request increased funding based on increased ridership, or to stretch limited funds farther, or a combination of both.
The goal of the optimization process is to increase efficiency of the transit network, with the inherent benefit of increasing ridership. This goal is achieved in several ways. Primarily, the optimization system collects information pertaining to all of the above-listed considerations and generates transit routes for the transit network along with vehicle and driver assignments for these routes. Each consideration is assigned a weight by the party seeking to optimize the network. For example, a priority could be set to maximize inter-regional travel, resulting in increased commuter train service and more inter-regional bus routes, while eliminating other regional bus routes and reducing inter-regional fare surcharges. In general, the result should provide an optimal combination of vehicle capacity, regional service and financial balance.
Another use for the optimization process is to provide an optimized destination-to-destination trip itinerary for any individual passenger subject to any specific limitations requested by that passenger. This way, passengers who want to use the transit network are readily provided with the information necessary to make a conscious and informed decision on how to use the transit network. The optimal itinerary generated can be based on priorities such as lowest fare, fewest transfers or closest transit stops to departure point and/or destination point.
Lastly, the combination of the three considerations set forth above should be continually reviewed to determine which changes must be implemented to achieve the desired goal. Possible changes include fare prices, addition/subtraction of vehicles and the addition/subtraction of routes. Also, the optimization data can be used for future planning, such as building new subway/train stations and extending service to new geographic regions. Finally, optimization allows for improved tracking of riders, which can be used to adjust the funding levels received from government and charitable sources.
For a paratransit network as either part of the whole transit network or as a separate network the optimization process works in the same way. Additionally, with paratransit, the number of riders serviced may be increased by changing the number and types of vehicles and routes available. This may be done by incorporating existing public transit services into the paratransit services where possible.
As shown in
By implementing the optimization system as shown in
Bus routes and schedules are received from the bus network, train stops and schedules from the train network, and so forth. Notably, information from other networks is readily incorporated into the whole. Paratransit networks can enter vehicle lists, availability and passenger eligibility criteria. Commercial transit networks, such as airlines, can enter flight schedules or other relevant information. By uniting all this data in one location, a uniformity of content can be provided, which enables uniformity of results when data inquiries are made.
Besides the inherent advantages the system provides for a passenger, there is also a large advantage gained by the transit operators. The accumulation of data allows for greater data analysis and data mining to improve the efficiency of the transit services provided.
At this stage, all of the optimization considerations are set forth. The different types of vehicles and routes available must be considered. Typical optimization factors include overlap between bus route and subway routes, connection times for train arrivals and corresponding bus departures, identification of transfer points where riders must switch vehicles.
Regional issues must be factored in, including the identification of regional boundary zones, specific identification of inter-regional routes and targeting of “hubs”, such as central train stations and subway station, for specific optimization.
Significantly, all those considerations act in concert with one another. For example, inter-regional travel (a geographic consideration) may require an additional fare (a financial consideration) depending on which vehicle and route are used (a routing consideration). If the rider takes an inter-regional commuter train, the additional fare may be built in to the base fare price, although commuter trains typically operate in limited time periods, so additional bus service between regions may also be provided. The bus service may also include passengers who are not traveling between regions, so separate types of fares may be needed.
The optimization process looks at these different considerations and provides results that take all of them into account. In this example, one result might be to reduce the additional inter-regional fare to increase ridership. Alternatively, adding an additional commuter train could reduce demand for inter-regional bus service, allowing those vehicles to be reassigned to other existing or new routes. Another possibility is to modify the boundaries of the regions to increase the scope of regional service. Or all of these optimization suggestions may be combined. These results are derived from incorporating all relevant considerations into the optimization process.
By examining a large range of data, transit operators are able to build a fairly accurate picture of the future. Trends can be incorporated, such as population density decreasing in downtown areas as jobs move further and further into the suburbs; aging populations increasingly dependent on traditional or paratransit services; declining road systems that suffer from near-constant gridlock; etc. Thus, operators can explore the role transit services might play in better adapting to meet the dynamic needs of the public using the subject invention. Demand forecasting and spatial analysis can be used to demonstrate the effects of new bus routes, expanded demand response delivery, route-deviated services, light rail networks or different fare structures.
The optimization process can also help predict which types of service changes will have the most positive affects on ridership. The invention can evaluate proposed routes against a number of criteria. In procuring financial and policy support from all levels of government, transit operators must be able to present a compelling look ahead, and clearly articulate their strategies for dealing with it. The optimization system provides mechanisms for forecasts, analysis and scenario modeling, which in turn allow staff and consultants to spend their time working on the solutions rather than searching for the problem.
An example of the optimization process is readily demonstrated by its application to passenger scheduling and itineraries. The information gathering and processing performed by the optimization process adds several advantages to passenger scheduling that would not otherwise be available. The primary result is that a single passenger scheduling algorithm can be used where previously multiple passenger scheduling algorithms were required
One advantage is that a single algorithm can now be used to handle multiple modes of transport. A user enters a trip request (consisting of origin, destination and desired departure time) and asks the algorithm to suggest possible transit solutions. The algorithm returns a list of possible itineraries that includes paratransit vehicles, fixed route vehicles, flex route vehicles and even taxis and connections to other PTOs and CTOs. With the prior art using different algorithms for different operators, the trip request would have to be directed to each algorithm independently, in some cases requiring that the trip request data be formatted differently. Also, each algorithm would return the results in a different data format, placing a burden on the user interface to try to integrate them all.
The biggest separation between the past algorithms was between the fixed route and variable-route (paratransit) algorithms. Consider using a city's existing itinerary lookup website and requesting directions on how to get from point A to point B. It might tell you to get on a certain bus at route 101, transfer to a different bus at route 400 and so on. With this optimization process behind it, the algorithm can now also suggest potential alternatives such as taking a taxi or a dial-a-ride vehicle instead of, or in combination with, using the fixed route buses. Existing transit algorithms do not provide that degree of integration. This algorithm advances beyond this limitation through the use of the optimization system.
Another advantage arises in the algorithm's ability to generate transfers between vehicles of different types. Not only can it return solutions where each solution uses a different transit type, but it can also return a solution that combines multiple types into one single itinerary. For example, it might suggest a solution where a passenger is picked up by a paratransit bus, taken to a transfer location where they transfer to a fixed route bus, then get off at another transfer point where they transfer to a taxi for the last leg of the trip. This type of solution offers a passenger more choices. For instance, it can allow a passenger to book trips to places where fixed route buses are unavailable while still using the fixed route bus for a portion of the trip. It can also save money because fixed route buses are usually cheaper to use than paratransit buses or taxis. For example, a special needs passenger that cannot walk to a bus stop would traditionally have to take a paratransit bus or taxi from door to door. This service is typically provided by paratransit operators who are heavily subsidized by taxpayers. On long trips, if the paratransit operators can use fixed route buses for part of the trip then they can potentially save money. Backed by the information gathered by the optimization system, the algorithm allows a variable route such as paratransit or taxi to pick the passenger up right where they are and also to drop them off right where they need to go while still using fixed route for portions of the trip in between.
A third advantage gained by using the algorithm is that it unifies the solution costing model across all the types of transport. A solution cost is an abstract number assigned to each potential solution. It is used by the transit operators to help judge which solution is the best choice. It can be based on many factors such as the amount of extra distance added to the vehicles, the number of transfers involved, vehicle load utilization, passenger on-board time and many others. With separate algorithms it is difficult to compare costs for the same trip request because each algorithm has its own way of digesting the multiple cost factors into one single cost. By popular analogy, it was like comparing apples to oranges. However, with all the solutions generated by a single algorithm, it is possible to unify the costing model so that the relative costs of solutions involving different transport modes can be fairly compared, enabling more intelligent selections to be made. To complete the analogy, using the algorithm, comparing a paratransit solution to a fixed route solution is like comparing apples to apples.
In
The passenger scheduling algorithm works by dividing the trip solution generation process into several phases. Phase 1 is the discovery of which transit services operate in the vicinity of the trip's origin and destination points. Phase 2 is the discovery of transfer patterns between origin and destination transit services. Phase 3 involves finding times and vehicles for each segment of each transfer pattern found in phase 2. Phase 4 calculates the cost of the valid solutions found in phase 3.
Phase 1 introduces a concept called a transit service. A transit service is an abstraction of all the types of transport that the algorithm supports. A service combines a type of transport with a representation of the area serviced by that type, in effect a combination of the vehicle and route data as well as the geographic and demographic data that has been gathered and analyzed by the optimization system. How the area is defined depends on the type of transport. For a fixed route vehicle, it is defined by the bus stop pattern. A fixed route vehicle always follows a particular path through the city streets. That path can be represented by its bus stop locations connected together in a line. A paratransit vehicle has no fixed stops to define it. Its route is ultimately defined by the trips it is assigned each day and that can vary from day to day. In addition, many of the trips are not assigned until shortly before it pulls out for the day. As a result, the trips on a paratransit bus cannot be used to define a paratransit service because they are not a fixed entity. Like a taxi, a paratransit bus can go anywhere but for practical purposes it is often restricted to certain regions, or zones, of the city. Paratransit and taxi services can therefore be better represented by a polygon that defines their service area rather than by a sequence of pre-determined fixed stops. The passenger scheduling algorithm allows a variable route service to be defined either way—either using the fixed stops assigned to the variable route, or by using a service area polygon assigned to the variable route. This is because a variable route is fundamentally considered as a hybrid between a fixed route bus and paratransit bus.
Phase 1 involves finding all the transit services that operate in the area of the passenger's origin plus all the transit services that operate in the area of the passenger's destination. The services are then sorted, preferably in order of proximity to the actual passenger's location. An alternative way to look at this is that they are sorted in order of how far the passenger would have to walk to use each service. For some modes, like paratransit or taxi, the walking distance is essentially zero because these modes go directly to the passenger's origin or destination point. For services based on stop points the proximity is measured as the walking distance to the nearest bus stop for that service. For services based on polygons, the proximity is based on whether or not the passenger's origin or destination is contained within that polygon. This phase produces a set of from/to service pairs but makes no attempt to choose a vehicle or departure times or to work out a transfer pattern between the services. The transfer pattern is left to phase 2.
Phase 2 is the center of the passenger scheduling algorithm. This phase is where the different types of transport are integrated. Phase 2 accepts the list of from/to services that were produced by phase 1 and then works out transfer patterns between each of those service pairs. Phase 2 uses a recursive algorithm to walk through a transfer table. It can work out all possible ways of transferring between the origin and destination services. The transfer table is set up in advance as part of the optimization process. It allows the passenger to decide which locations are good for making transfers and which services to include or exclude from any particular transfer. However, users do not have to define the complete transfer patterns—only the transfers between two adjacent services. The algorithm in phase 2 then does the rest by dynamically building more complex transfer patterns out of the simple from/to transfer pairs.
In the present example, phase 2 generates coded transfer patterns based on a transfer table. For example, the pattern A-1-E-3-B means: “take service A to stop 1 then transfer to service E which you take to stop 3 where you transfer to service B which takes you to the final destination.” This particular transfer pattern is illustrated by the highlighted lines in
In Phase 3, each pattern must be expanded into a full solution that involves specific vehicles and arrival and departure times. This involves starting with the first service in the pattern and finding all possible vehicles for that service that can pick up the passenger and take them to the next drop-off point. If there is a next service, then the drop-off point will be a bus stop where the transfer takes place and so on down the line. When there is no next service then the drop-off point will be the final destination. The pickup and drop-off times are worked out for each vehicle. When moving on to the next service in the pattern, the pickup time is restricted based on a small transfer window around the previous drop-off. For example, if a vehicle drops a passenger off at a transfer point at 9:00 am, then the vehicle used by the next service to pick them up must do so within the window of 9:00 am to 9:20 am (assuming a maximum allowed layover time of 20 minutes). This window helps to restrict the possible vehicle choices for the services involved in picking up a transfer passenger. The travel time to the stop is then calculated which determines the next drop-off time and so on down the line until the end of the pattern is reached. Once all the services in a pattern have their vehicles and times worked out, then it is becomes a solution and gets added to a list of valid solutions. Sometimes a pattern cannot produce a solution because it does not have any vehicles available at the appropriate time window. In this case the pattern is thrown out. It should be noted that a single transfer pattern can generate multiple solutions because each service can offer multiple vehicles and times to choose from.
Phase 4, the last phase in the passenger scheduling algorithm, accepts the list of valid vehicle/time solutions and then calculates the relative cost of each one. This is where the universal costing formula is applied to all the vehicle types in each solution. The end result of phase 4 is that the list of valid solutions is sorted in order of ascending cost and then presented to the user for selection. The user may choose to select the lowest cost solution or they may choose to re-sort the list based on different criteria i.e. fewest transfers.
The route planning selection described above can be similarly applied by transit operators when dispatching vehicles and drivers on routes. For fixed routes, it is a relatively straightforward process to assign vehicles and drivers to cover the routes to minimize deadheading and comply with any necessary labor regulations. However, for variable routes, optimization of dispatching is much more difficult. By reviewing the most common requests, both routes and times, for a variable route service, the optimization process may provide alternative solutions for dispatching. For example, larger or smaller capacity vehicles may be used, or additional vehicles added to a route to accommodate the passenger traffic on a route with a minimum of wasted space. Shift changes and break times for drivers can be adjusted to reflect slack periods in service demand.
One use of the optimization process is to reduce personnel expenditures such as overtime by allowing the transit operators to dynamically alter schedules to ensure that each route is covered. Another is the ability to monitor on-board vehicle systems, such as idle monitors, which can also be used as part of driver evaluation processes.
An additional consideration is that drivers can readily access the route system, similarly to passengers, allowing driver input to be more quickly incorporated into the optimization process. Furthermore, driver input can be solicited as a valuable addition to the optimization system's projections by adding driver observations about items such as traffic density, stop locations and similar aspects of the system that can benefit from observer information.
The passenger information services provided by the optimization system also needs to accommodate new technologies that will, over the long term, need to interface with the system. For example, most information services will soon need to be accessible not just through call centers or PCs, but also to web-enabled cell phones, handheld computers and PDAs, interactive television, or an automated voice response system. The optimization system must allow users to enter not only dates and times of travel and to choose departure locations and destinations by street address intersection, but also to choose locations from a list of common locations such as hospitals, shopping centers or tourist attractions. Other services and priority settings available include: minimizing walking distances, minimizing number of transfers, minimizing travel times, identifying preferred travel mode, identifying ADA routes, clicking on a map to determine departure and arrival locations.
Output options include a basic trip summary with travel time, distance and fares, a detailed written itinerary or even a map with origin, destination and transfer points clearly marked. The system also provides for: return trip planning, street routing, detailed walking instructions, multi-lingual services, multimodal travel information, next bus information, real-time schedule information, and accessibility for passengers with physical or cognitive limitations.
This concludes the description of a presently preferred embodiment of the invention. The foregoing description has been presented for the purpose of illustration and is not intended to be exhaustive or to limit the invention to the precise form disclosed. Many modifications and variations are possible in light of the above teaching and will be apparent to those skilled in the art. It is intended the scope of the invention be limited not by this description but by the claims that follow.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2,554,651 | Jul 2006 | CA | national |
