Real-time incorporation of new orders into ongoing VRP optimization processes
The described system addresses the challenge of incorporating real-time order inputs into VRP optimization by processing new orders through a network system that calculates and integrates new slots and destinations, enhancing operational efficiency and practicality in vehicle routing.
Patent Information
- Application Number
- PCT/US2024/028370
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2023-05-15
- Filing Date
- 2024-05-08
- Publication Date
- 2024-11-21
AI Technical Summary
Current vehicle routing problem (VRP) optimization systems face challenges in incorporating real-time order inputs, such as new deliveries and destinations, due to their NP-hard complexity, requiring significant computational resources and time, which is impractical for commercial applications, especially when dealing with capacitated VRP with time windows (CVRP with TW).
A computer-based network system that includes an order processing platform, real-time incorporation platform, VRP optimization platform, and batching and dispatch platform, which processes new orders in real-time by calculating permutations of available slots, determining compatible slots, and integrating new payloads and destinations into ongoing VRP optimization processes, minimizing disruption to existing solutions through repeated operations of the VRP solver.
Enables real-time incorporation of new orders into ongoing VRP optimization processes, reducing computational costs and maintaining efficient vehicle routing solutions, thereby improving operational efficiency and practicality for real-world applications.
Smart Images

Figure US2024028370_21112024_PF_FP_ABST
Abstract
Description
REAL-TIME INCORPORATION OF NEW ORDERS INTO ONGOING VRPOPTIMIZATION PROCESSESTECHNICAL FIELD
[0001] The present application relates to apparatuses, methods, systems, and techniques for real-time incorporation of new orders into ongoing vehicle routing problem (VRP) optimization processes.BACKGROUND
[0002] A number of proposals have been made for determining optimized solutions of the vehicle routing problem (VRP). Since the VRP has non-determini Stic polynomial time hard (NP- hard) complexity, finding an optimized solution is impossible within commercial time frames even using extraordinary computing resources. Thus, all practical solutions must use approximations which typically focus on finding a feasible solution that approximately minimizes the total distance traveled by the fleet, the total travel time of the fleet, or other cost function metrics as a primary goal. Even using such approximations, the time and computing power required for determining optimized VRP solutions is beyond the capability of human calculation and requires significant computer resources and computing time. While VRP optimizations are technically challenging, trying to accommodate real-time order inputs, such as the addition of deliveries including new payloads and new destinations and providing delivery time selection of such additions, and computation required to determine possible slots for added deliveries (e.g., time and / or space slots), poses even greater technical challenges. Accommodating and incorporating such inputs in VRP optimization systems and processes requires a more complicated approach, such capacitated VRP with time windows (CVRP with TW) and poses a substantial unresolved technical challenge to practical, real-world adoption and utilization. There remains a significant unmet need for the unique apparatuses, processes, systems, and technical solutions of the present disclosure.DISCLOSURE OF EXAMPLE EMBODIMENTS
[0003] For the purposes of clearly, concisely, and exactly describing example embodiments of the present disclosure, the manner, and process of making and using the same, and to enable the practice, making and use of the same, reference will now be made to certain example embodiments, including those illustrated in the figures, and specific language will be used to describe the same. It shall nevertheless be understood that no limitation of the scope of the invention is thereby created, and that the invention includes and protects such alterations, modifications, and further applications of the example embodiments as would occur to one skilled in the art.SUMMARY OF THE DISCLOSURE
[0004] Some embodiments according to the present disclosure include unique apparatuses, methods, processes, and techniques for real-time incorporation of added payloads and destinations into ongoing vehicle routing problem (VRP) optimization processes. Further embodiments, forms, objects, features, advantages, aspects, and benefits shall become apparent from the following description and drawings.BRIEF DESCRIPTION OF THE DRAWINGS
[0005] Fig. l is a schematic diagram illustrating certain aspects of an example vehicle delivery route planning and dispatch network.
[0006] Fig. 2 is a flow diagram illustrating certain aspects of an example process executable in connection with the network of Fig. 1.
[0007] Fig. 3 is a flow diagram illustrating certain aspects of an example process executable in connection with the network of Fig. 1 or the process of Fig. 2.
[0008] Fig. 4 is a flow diagram illustrating certain aspects of an example process executable in connection with the network of Fig. 1.
[0009] Fig. 5 is a flow diagram illustrating certain aspects of an example process executable in connection with the network of Fig. 1 or the process of Fig. 3.DETAILED DESCRIPTION OF EXAMPLE EMBODIMENTS
[0010] With reference to Fig. 1, there is illustrated an example vehicle delivery route planning and dispatch network 100 (also referred as network 100). Network 100 is one example of a computer-based network configured for and operable to provide multiple options in real-time incorporation of new deliverable payloads and destinations into an ongoing VRP optimization process. According to the present disclosure a deliverable payload (sometimes referred to herein simply as a “deliverable” or a “payload”) may refer to an item or object to be delivered to a destination or a human rider who will provide a service at a destination. Deliverable payloads according to various embodiments of the present disclosure may be represented to by data structures or objects comprising a number of attributes and parameters including, for example, physical characteristics such as mass, volume, weight, temperature requirements and others physical characteristics, temporal characteristics such as transaction duration (e.g., in the case of a service to be rendered or an item or object with delivery time requirements to be accounted or), as well as other characteristics.
[0011] In the illustrated example, network 100 comprises a plurality of platforms, namely order processing platform 110 (also referred to as platform 110), real-time incorporation platform 130 (also referred to as platform 130), vehicle routing problem (VRP) optimization platform 150 (also referred to as platform 150), and batching and dispatch platform 170 (also referred to as platform 170). The foregoing and other platforms of network 100 may be implemented in one or more computing systems which may be implemented and provided in a number of forms including, for example, as one or more cloud-based computing systems, data center-based computing systems, stand-alone computing systems, distributed or virtual computing systems, or combinations thereof.
[0012] Platform 110 includes an ordering interface which is accessible by and configured to receive orders from computer systems external to network 100, for example, from a customer computer system, an intermediary computer system, or other third-party computer system. Platform 110 is further configured to process orders received from such external computer systems and to provide information pertaining to such orders to platform 130. In the illustrated example, platform 110 is depicted as providing real-time incorporation (RTI) order information 112 to platform 130. Platform 110 may provide RTI order information 112 to platform 130 in a numberof manners including, for example, by transmitting to, linking to, permitting access to, or otherwise communicating RTI order information to platform 130.
[0013] RTI order information 112 may be provided by platform 110 to platform 130 in a number of forms. In certain instances and in connection with certain operations, RTI order information 112 may comprise a new order reservation (NOR) including a new payload and destination of an order received by platform 110. The NOR may be generated in response to and may include information relating to an order that has been initiated but not yet completed or consummated. In certain instances and in connection with certain operations, RTI order information 112 may comprise new order confirmation (NOC) including a new payload and destination of an order received by platform 110. The NOC may be generated in response to and may include information relating to an order that has been completed or consummated.
[0014] Platform 150 includes optimization parameters 149, VRP solver 152, VRP optimization solutions 154, optimization snapshot interface 156, and solution update interface 158. VRP solver 152 is configured to determine an optimized solution of a non-deterministic polynomial-time hard (NP-hard) vehicle routing problem (VRP) subject to optimization parameters 149.
[0015] Optimization parameters 149 may be based in whole or in part upon information provided from platform 130 which may, in turn, be based on information received from platform 150. For example, optimization parameters and / or optimization constraints 147 may be generated, modified, and / or otherwise influenced or informed by components or processes of platform 130 (e. ., outputs or parameters of slot calculator 134, compatibility evaluator 136, and / or other components or processes of platform 130) and, in turn, the components or process of platform 130 may be generated, modified, and / or otherwise influenced or informed by optimization snapshot interface 156 or other components or processes of platform 150. In some embodiments, optimization parameters and / or optimization constraints 147 may be generated, modified, and / or otherwise influenced or informed by optimization snapshot interface 156 which may be provided directly to optimization parameters 149 or provided via platform 130.
[0016] Optimization parameters 149 include optimization objectives 143, optimization parameters 145, and optimization constraints 147. The optimization parameters typically include a set of destinations (e.g., delivery destinations and / or pickup destinations), one or more payloads associated with each of the destinations, a set of vehicles available for the optimization, and a setof drivers available for the optimization. Tn some embodiments, the optimization parameters may include additional or alternative parameters to be utilized in the optimization.
[0017] The optimization constraints include a set of constraints imposed on the VRP optimization solution determined by VRP solver 152. The optimization constraints may include hard constraints that the VRP solution is required to satisfy as well as soft constraints that the VRP solution is requested to satisfy, but which may tolerate excursions from the constraints. The optimization constraints for a payload comprising an object of item to be delivered may include payload volume constraints and payload mass or weight constraints for a given vehicle. The optimization constraints for a payload comprising a rider who will provide services may include Boolean resource and skills parameters, for example, whether a given vehicle has refrigeration or whether a given rider has certain resources or skills associated with or required by one or more services. The payload volume constraints and payload mass or weight constraints for a given vehicle may be hard constraints, but may also be implemented as soft constraints, for example, by defining a constraint less than an actual physical constraint of the vehicle and permitting excursion up to the actual physical constraint or up to some other level up to the actual physical constraint. In some embodiments, the optimization constraints may include additional constraints including, for example zoning restrictions, route restrictions. For example, a set of one or more vehicles may be associated with a service region bounded by a polygon on a map space or domain to define an operating zone. Additionally or alternatively, route restrictions may be imposed to restriction or limit operation of a set of one or more vehicles on one or more routes at one or more times (e.g., trailers may be excluded from city traffic routes during the day or during high traffic times). In some embodiments, the optimization constraints may include additional constraints including, for example, driver shift or working hour limits, vehicle service duration constraints, and / or other constraints to be imposed on the optimization.
[0018] The optimization objectives include one or more objectives for the VRP optimization solution to be determined by VRP solver 152. The optimization objectives may be provided in the form of an objective function. In some embodiments, the optimization objectives may seek to minimize distance traveled. In some embodiments, the optimization objectives may seek to minimize additional or alternative objective function metrics for completing optimization parameters, such as fuel consumption metrics, energy usage metrics, or other objective function metrics.
[0019] Platform 150 is configured to repeatedly operate VRP solver 152 to determine an optimized solution of an NP-hard VRP. The optimized solution may define a plurality of routes for a plurality of vehicles. Each of the routes may include a plurality of destinations for a respective vehicle of the plurality of vehicles and a respective payload for each of the plurality of destinations. Through repeated operation of VRP solver 152, a plurality of VRP optimization solutions 154 may be determined. The plurality of VRP optimization solutions 154 may include solutions for multiple days (denoted as DI, D2, ... Dn in the illustrated example). The plurality of VRP optimization solutions 154 may also include multiple solutions for a given day (denoted as SI, S2, ... Sn in the illustrated example). At any arbitrary point in time, various numbers of the plurality of VRP optimization solutions 154 may be extant and stored in a non-transitory memory medium of or associated with platform 150. VRP solver 152 may utilize a selection of an available slot determined by platform 130 (e.g., a customer selection of an available slot determined or offered by platform 130) as a basis for determining subsequent optimized solutions of the NP-hard VRP. This allows for the incorporation and preservation of an existing solution so that that disruption of an existing solution is mitigated, minimized, or eliminated. VRP solver 152 may also utilize a snapshot received via optimization snapshot interface 156 as a basis for subsequent solutions and may receive or utilize such snapshot information, for example, via platform 130 or directly to optimization parameters 149.
[0020] In the illustrated embodiment, platform 130 includes a plurality components including snapshot capturer 132, slot calculator 134, compatibility evaluator 136, slotter 138, and batching selector 139 which may be configured to implement and execute one or more operations of a process such as process 200 described below in connection with Fig. 2. In other embodiments, platform 130 may include additional or alternative components including, for example and components that combine the features and / or function of two or more the components depicted in the illustrated example, and / or multiple components that include divided portions of the features and / or function of the components depicted in the illustrated example.
[0021] Snapshot capturer 132 is configured and operable to obtain a snapshot of a current state of at least one of VRP optimization solutions 154 determined VRP solver 152. Snapshot capturer 132 may interface with optimization snapshot interface 156 of platform 150 and may request or otherwise obtain a snapshot of a current state of at least one of VRP optimization solutions 154 determined VRP solver 152 from or via optimization snapshot interface 156.
[0022] Slot calculator 134 is configured and operable to calculate a set of permutations of available slots in the snapshot for the new payload and destination. In some embodiments, the set of permutations may include all potential permutations. In some embodiments, the set of permutations may include a subset of all potential permutations effective to decrease computational cost relative to all potential permutations. The available slots may be of a relatively narrow predetermined range (e.g., 1-2 hours, or 1-4 hours, or some other fraction of a day or vehicle operating shift). Additionally or alternatively, the available slots may comprise expanded predetermined ranges, for example, a 4 to 6 hour window rather than a one to two hour window to facilitate offerings of slots to customers who may be willing or incentivized to accepts larger delivery time range which adds an additional degree of freedom to the optimization.
[0023] Slot calculator 134 may be configured to calculate a set of permutations of slots available in the snapshot for the new payload and destination across multiple days. Slot calculator 134 may also be configured to calculate a set of permutations of slots available in the snapshot for the new payload and destination using operations and techniques such as those described below in connection with Figs. 2 and 3.
[0024] Compatibility evaluator 136 is configured to determine a compatible slot of the set conforming to one or more optimization constraints applicable to the snapshot. Compatibility evaluator 136 may be configured to obtain or receive from platform 150 one or more optimization constraints 147 of optimizations parameters 148. Compatibility evaluator 136 may be configured to determine a compatible slot of the set conforming to one or more optimization constraints applicable to the snapshot using operations and techniques such as those described below in connection with Figs. 2 and 3.
[0025] Slotter 138 is configured to provisionally add the new payload and destination of the NOR to one or more compatible slots of the snapshot and to optimization parameters 149 utilized by the VRP solver 152. After being added to the optimization parameters 149, the new payload and destination may be utilized by the VRP solver 152 in determining additional optimizations provided to VRP optimization solutions 154. Slotter 138 may be configured to provisionally add the new payload and destination of the NOR to the compatible slot of the snapshot and to optimization parameters utilized by the solver using operations and techniques such as those described below in connection with Figs. 2 and 3. Such provisional addition or assignment may, for example, be subsequently confirmed, selected, and / or validated by a customer or other enduser. Such operations may occur, for example, at an external customer system or a customer interface of order processing platform 110 such as a web portal or API interface or another system or interface of system 100.
[0026] Batching selector 139 is configured to select for batching and dispatch one of (a) the snapshot of the current VRP optimization with the new payload and destination of the NOR added to the compatible slot by slotter 138 and (b) a new optimized solution determined by VRP solver 156 with the new payload and destination of the NOR included as optimization parameters 154. This selection may be made, for example, based on a current state of VRP solver such as VRP solver 156 which may be provide by an associate state machine and / or based on addition of a new optimization solution to optimization solutions 154 which may be communicated by or received from solution update interface 158. Batching selector 139 may be configured to perform these actions in response to an order confirmation received from the order processing platform after the order reservation has been received from the order processing platform. The new optimized solution determined by VRP solver 156 with the new payload and destination included as optimization parameters may be obtained or received by batching slotter 138 from or via solution update interface 158 or platform 150.
[0027] Platform 170 is configured to batch a vehicle dispatch plan using the selected one of (a) the snapshot of the current VRP optimization with the new payload and destination of the NOR added to the compatible slot by slotter 138 and (b) a new optimized solution determined by VRP solver 156 with the new payload and destination of the NOR included as optimization parameters 154. The batch dispatch plan is provided as dispatch outputs to a plurality of vehicles to control or determine operation, payloads, destinations and routes of the plurality of vehicles.
[0028] With reference to Fig. 2, there is illustrated an example process 200. Process 200 is one example of a computer-implemented process configured for and operable to provide real-time incorporation of new payloads and destinations into an ongoing VRP optimization process.
[0029] In the illustrated example, process 200 begins at start operation 202 and proceeds to operation 204 which associates process 200 with a vehicle routing problem (VRP) solver, such as VRP solver 152 of system 100. The VRP solver may perform a process 290 wherein the VRP solver is repeatedly operated at operation 292 to determine an optimized solution 294 of a non- deterministic polynomial-time hard vehicle routing problem as indicated by the looped relationship of operation 292 and optimized solution 294. Each iteration of optimized solution 294 may definea plurality of routes for a plurality of vehicles. Each of the routes may include a plurality of destinations for a respective vehicle of the plurality of vehicles and a respective payload for each of the plurality of destinations. One or more iterations of optimized solution 294 may be included and stored as a plurality of optimized solutions such as optimized solutions 154 of system 100.
[0030] The association of process 200 may with a VRP solver may comprise a number of operations. In some embodiments, the association may be preestablished, for example, by an association between a system performing process 200 and a system operating the VRP solver. In some embodiments, the association may include checking or evaluating the operation or operational state of the VRP solver. The checking or evaluating may evaluate whether the operation or operational state of the VRP solver is compatible with process 200, for example, whether operation or state of the VRP solver can service or provide information needed for one or more of the operations of process 200.
[0031] At a point during the repeatedly operating the VRP solver, process 200 proceeds to operation 206 which monitors for a new order reservation (NOR) which may be provided by and received from an order processing platform such as order processing platform 110 of system 100. It shall be appreciated that the NOR is one example of an order reservation including a new payload and destination, the new payload and destination not being included in any optimized solution determined by the solver at the current point during the repeatedly operating the solver.
[0032] From operation 206, process 200 proceeds to conditional 208 which evaluates whether a NOR is received. If conditional 208 evaluates negative, process 200 proceeds to operation 206 and further proceeds as described above. If conditional 208 evaluates affirmative, process 200 proceeds to operation 210. It shall be appreciated that operation 206 and conditional 208 provide one example of receiving an order reservation from an order processing platform.
[0033] Operation 210 obtains a snapshot of a current state of the optimized solution of the VRP solver. The snapshot may include optimized solution may comprise the selecting a set of optimizations including the most recent optimizations completed by the VRP solver for one or more days including for multiple days (or one or more other time periods such as vehicle shifts). With reference to the embodiment of Fig. 1, at a given point of time the VRP optimization solutions 154 may include a set of optimizations including the most recent optimizations completed by the VRP solver 156. For example, at one such point in time the set may include solutions D1-S2 andD2-S1 or any of various other sets of the solutions depicted in the illustrated example. From operation 210, process 200 proceeds to operation 212.
[0034] Operation 212 calculates a set of permutations of slots available in the snapshot for the new payload and destination. Operation 212 may be configured to calculate a set of permutations of slots available in the snapshot for the new payload and destination by identifying each available vehicle plan in a snapshot of a current optimization that is present or operating for each defined slot on each defined day for slotting, calculating the closest vehicle plans to the NOR, calculating a set of all permutations for adding the NOR to the set of closest available vehicle plans, and calculating those permutations of the set including serviceable payload capacity. The calculation of permutations of available slots may also account for capacity constraints and resource constraints, such as capacity constraints on specific vehicle capabilities, for example, the amount of refrigerated space available in a vehicle as opposed to the total space available. Operation 212 may be further configured to calculate a set of permutations of slots available in the snapshot for the new payload and destination using operations and techniques such as those described in connection with Fig. 3. From operation 212, process 200 proceeds to operation 214.
[0035] Operation 214 determines one or more compatible slots of the set satisfying or otherwise conforming to one or more optimization constraints applicable to the snapshot. From operation 214, process 200 proceeds to conditional 216 which evaluates whether one or more compatible slots were determined by operation 214.
[0036] If conditional 216 evaluates negative, process 200 proceeds to operation 219 which requests a re-solve operation and, in some embodiments and some instances may adjust one or more soft constraints of the constraints utilized by operation 214. From operation 219, process 200 proceeds to operation 210 and further proceeds as described above.
[0037] If conditional 216 evaluates affirmative, process 200 proceeds to operation 218 which provisionally adds the new payload and destination of the NOR to one or more comparable compatible slots of the snapshot and to optimization parameters utilized by the solver. From operation 218 process 200, proceeds to operation 220. Such provisional addition or assignment may, for example, be subsequently confirmed, selected, and / or validated by a customer or other end user.
[0038] Operation 220 monitors for a new order confirmation (NOC) which may be provided by and received from an order processing platform such as order processing platform 110 of system100. It shall be appreciated that the NOC is one example of an order confirmation including an indication confirming the new payload and destination and an indication for including in a batching and dispatch plan for servicing the indicated payload and destination.
[0039] From operation 220, process 200 proceeds to conditional 222 which evaluates whether a NOC was identified by operation 220. If conditional 222 evaluates negative, process 200 proceeds to conditional 223 which evaluates a timeout condition. If conditions 223 evaluates negative, process 200 proceeds to operation 220 and further proceeds as described above. If conditional 223 evaluates affirmative, process 200 proceeds to end operation 299 and may thereafter be repeated.
[0040] If conditional 222 evaluates affirmative, process 200 proceeds to operation 224 which performs a batching selection. The batching selection performed by operation 224 includes selecting for batching and dispatch one of (a) the snapshot with the new payload and destination added to the compatible slot and (b) a new optimized solution determined by the solver with the new payload and destination included as optimization parameters. This selection may be made, for example, based on a current state of VRP solver such as VRP solver 156 which may be provided by an associate state machine and / or based on addition of a new optimization solution to a plurality of optimization solutions.
[0041] From operation 224, process 200 proceeds to operation 226 which performs a batching according to the selected one of the one of (a) the snapshot with the new payload and destination added to the compatible slot and (b) a new optimized solution determined by the solver with the new payload and destination included as optimization parameters and generates and provides a dispatching output to a plurality of vehicles to control or determine operation, payloads, destinations and routes of the plurality of vehicles. From operation 226, process 200 proceeds to end operation 299 and may thereafter be repeated.
[0042] With reference to Fig. 3, there is illustrated an example slot calculation and slotting process 300 (also referred to herein as process 300). Process 300 is one example of a process configured and executable perform slot calculation, compatibility determination, and assignment operations in connection with a computer system-based system or process providing real-time incorporation of new payloads and destinations into ongoing VRP optimization processes such as system 100 described above in connection with Fig. 1 and / or process 200 described above in connection with Fig. 2.
[0043] Process 300 begins at operation 302 and proceeds to operation 304 which obtains a set of slots and days which are to be evaluated in connection with a new order reservation (NOR) and / or a now order confirmation (NOC). The slots obtained and utilized by operation 304 may comprise predetermined time windows over a range of service, for example, a set of slots on a given day may comprise a slot from 8am to 10am, a slot from 10am to 12pm, a slot from 12pm to 2pm, and a slot from 2pm to 4pm. The days obtained and utilized by operation 304 may comprise days determined using a heuristic relating to the NOR (e.g., a range of future days for delivery based on assumed order preferences or derived from historical order preferences). The days obtained and utilized by operation 304 may additionally or alternatively be indicated by the NOR itself, for example, as requested delivery dates / times or date / time ranges.
[0044] From operation 304, process 200 proceeds to operation 306 which identifies all of the available vehicle plans (routes and deliveries serviced by a vehicle) in a snapshot of an optimization determined by a VRP solver for all of the defined slots and defined days obtained at operation 304. The available vehicle plans may be determined by the presence of vehicle plans whose execution falls into one or more of the defined slots on a given day. In some embodiments and executions the available vehicle plans may be filtered or limited (e.g., pre-filtered or prelimited) to vehicle plans with available payload space in vehicle plans whose execution falls into one or more of the defined slots on a given day.
[0045] In the illustrated example, the available vehicle plans are determined for slots SI (8am- 10am), S2 (10am-12pm), S3 (12pm-2pm), and S4 (2pm-4pm) and days DI and D2. The available vehicle plans are denoted by vehicle VI in slot SI on day 1; vehicles VI, V3 in slot S3 on day 1; vehicle V4 in slot S4 on day 1; vehicle V2 in slot SI on day 2; vehicles V2, V4 in slot S2 on day 2, and vehicles VI, V2, V3 in slot S3 on day 2, with no vehicles being associated with slot S2 on day 1 or slot S4 on day 2. It shall be appreciated that other embodiments and instances of execution may include a greater of fewer number of slots, and a greater of fewer number of days, or both. Likewise, in the illustrated example, additional or alternative available vehicle plans may be determined as available various slots and days.
[0046] From operation 306, process 300 proceeds to operation 308 which calculates the available vehicle plans that are closest to the NOR. Operation 308 may determine the closest available vehicle plans by identifying one or more of the closest destination nodes from one or more available vehicle plans. The one more closest destination nodes may be defined as meetinga closeness ranking criterion (e.g., top 10 closest, or some other number of closest nodes) and / or a proximity metric defining a threshold for closest destination categorization. In some embodiments, the closeness ranking criterion and / or the proximity metric may be utilized to evaluate all nodes of all available vehicle plans for closeness, select the closest node to a NOR destination, optionally remove the vehicle plan associated with the closest node from the evaluation, and repeat the selection of the closest node to the NOR destination and the optional removal for the vehicle plans remaining for evaluation. For example, a closest destination node (according to a closeness ranking criterion) may be selected and evaluated. If the closest destination node selection can be accommodated the selection may be confirmed and utilized by subsequent operations of process 300. If the closest destination node selection cannot be accommodated a next closest destination node (according to a closeness ranking criterion) may be selected and evaluated. This process may be iterated multiple times until a set of closest vehicle available vehicle plans is defined, for example, iterative selection and evaluation may be repeated iteratively until a selection is confirmed or all possible selections are exhausted. This technique may be utilized to avoid the need to calculate, select, or use static or fixed service destination clusters. Thus, operation 308 may dynamically define a plurality of modified vehicle plans as new dynamically modified clusters which may abrogate previous cluster definitions. This technique may allow a cluster to be dynamically defined or modified as each new incoming order is received.
[0047] From operation 308, process 300 proceeds to operation 310 which calculates the permutations available to add the NOR in each of the closest available vehicle plans calculated in operation 308. The available permutations may be determined based at least in part on the availability of payload space in each of the closest available vehicle plans. The illustrated example depicts the permutations available to add the new destination of the NOR (DNOR) into the plan for vehicle VI and shift SI, the plan for vehicle V2 and shift S3, and the plan for vehicle V4 and shift S4. It shall be appreciated that additional permutations may be available as denoted by ellipses and that in other embodiments or executions different vehicle-shift pairs may be determined.
[0048] From operation 310, process 300 proceeds to operation 312 which determines the calculated permutations that satisfy one or more constraints imposed on the VRP optimization that produced the snapshot. The one or more constraints may be hard constraints which may be intolerant of excursion or soft constraints tolerating excursion.
[0049] From operation 312, process 300 proceeds to operation 314 which provisionally adds the NOR to one or more satisfactory permutations. From operation 314, process 300 proceeds to conditional 316 which evaluates whether a new order criterion is obtained before a timeout condition occurs. If conditional 316 evaluates affirmative, process 300 proceeds to operation 318 which confirms the provisional assignment of operation 314. Operation 318 may, for example, provide the provisional assignment to a user or external system for confirmation, selection, and / or validation by customer or other end user. If conditional 316 evaluates negative, process 300 proceeds to operation 320 which discards or disregards the provisional assignment of operation 314. From either operation 318 or operation 320, process 300 proceeds to end operation 399 and may thereafter be repeated.
[0050] With reference to Fig. 4, there is illustrated an example process 400. Process 400 is one example of a computer-implemented process configured for and operable to provide real-time incorporation of new payloads and destinations into an ongoing VRP optimization process. Process 400 includes a number of operations and conditionals which are the same as or substantially similar to those illustrated and described in connection with process 200 and which are identified with the same reference numerals utilized in Fig. 2 and the description thereof.
[0051] Process 400 also varies from process 200 in certain respects. Process 400 is configured for and operable to provide real-time incorporation of new payloads into a cohort of vehicles that are already dispatched (also referred to herein as on-demand or in-dispatch incorporation), as opposed to providing such incorporation for cohorts of vehicles that are to be dispatched in the future such as in process 200. Accordingly, operation 210 obtains a snapshot of a current state of the optimized solution of the VRP solver in the form of a present state of an optimized solution corresponding to a dispatched cohort of vehicles. Additionally, process 400 omits operation 219 and does not request or attempt a re-solve operation which is inapplicable to a cohort of vehicle that have already been dispatched. Instead, if conditional 216 evaluates negative, process 200 proceeds to end operation 299. Furthermore, process 400 ma
[0052] Process 400 may be run concurrently with or in parallel with process 200 to provide both on-demand or in-dispatch incorporation of new payloads into a cohort of vehicles and similar incorporation into a cohort of vehicles to be dispatched in the future, such as on a subsequent day or shift. The incorporation solutions of process 200 and process 400 may be arbitrated to avoiddouble scheduling. For example, an incorporation solution by process 400 may be prioritized over and take precedence over an incorporation solution of process 200.
[0053] With reference to Fig. 5, there is illustrated an example slot calculation and slotting process 500 (also referred to herein as process 500). Process 500 is one example of a process configured and executable perform slot calculation, compatibility determination, and assignment operations in connection with a computer system-based system or process providing real-time incorporation of new payloads and destinations into ongoing VRP optimization processes such as system 100 described above in connection with Fig. 1 and / or process 400 described above in connection with Fig. 4. Process 500 includes a number of operations and conditionals which are the same as or substantially similar to those illustrated and described in connection with process 300 and which are identified with the same reference numerals utilized in Fig. 3 and the description thereof. Process 500 also varies from process 300 in certain respects.
[0054] Process 500 is configured for and operable to provide real-time incorporation of new payloads into a cohort of vehicles that are already dispatched (also referred to herein as on-demand or in-dispatch incorporation), as opposed to providing such incorporation for cohorts of vehicles that are to be dispatched in the future such as in process 300. Accordingly, operation 304’ generally corresponds to operation 304, but obtains a set of slots for a cohort of currently dispatched vehicles which are to be evaluated in connection with a NOR. Additionally, operation 306’ generally corresponds to operation 306, but identifies all of the available vehicle plans (routes and deliveries serviced by a vehicle) in a snapshot of an optimization determined by a VRP solver for the cohort of dispatched vehicles. Furthermore, from operation 314, process 500 proceeds to operation 315. Operation 315 is configured and operable to evaluate a slot demand and / or a slot difficulty, to score a plurality of slots according to slot demand and / or slot difficulty.
[0055] The slot demand evaluation and scoring of operation 315 may be based upon historical information about how in demand the slot is, for example, certain slots for certain missions such as plumbing or electrical services may be in high demand before or after typical business hours. The slot difficulty evaluation and scoring may be based upon one or more metrics of how much capacity or resource will be consumed by a given new order reservation (NOR), for example, by evaluating a distance or deviation from existing service locations, how long the new NOR would take, payload volume, or other metrics as will occur to one of skill in the art with the benefit and insight of the present disclosure. The scoring performed by operation 215 may be utilized toevaluate and plan service availability and pricing according to the slot demand and / or slot difficulty score(s), for example, by premium pricing for high demand, high difficulty and to more evenly distribute to improve utilization. Providing differential or premium pricing may incentivize and promote more complete or full utilization of available resources in terms of available slots, vehicles, service providers, or other assets.
[0056] As shown by this detailed description, the present disclosure contemplates multiple and various embodiments, including, without limitation, the following example embodiments. A first example embodiment is a process comprising: repeatedly operating a solver to determine an optimized solutions of a non-deterministic polynomial-time hard vehicle routing problem, the optimized solutions defining a plurality of routes for a plurality of vehicles, each of the routes including a plurality of destinations for a respective vehicle of the plurality of vehicles and a respective deliverable for each of the plurality of destinations; at a point during the repeatedly operating the solver, receiving an order reservation from an order processing platform, the order reservation indicating at least one of an item and a service and including a new deliverable and destination, the new deliverable and destination not being included in any optimized solutions determined by the solver at said point; obtaining a snapshot of a current state of the optimized solutions of the solver; calculating a set of permutations of slots available in the snapshot for the new deliverable and destination; determining a compatible slot of the set, the compatible slot conforming to one or more optimization constraints applicable to the snapshot; provisionally adding the new deliverable and destination to the compatible slot of the snapshot and to optimization parameters utilized by the solver; after receiving the order reservation, receiving an order confirmation from the order processing platform; and in response to the order confirmation, selecting for batching and dispatch one of (a) the snapshot with the new deliverable and destination added to the compatible slot and (b) a new optimized solutions determined by the solver with the new deliverable and destination included as optimization parameters; batching a vehicle dispatch plan using the selected one of (a) and (b); and dispatching a plurality of vehicles according to the vehicle dispatch plan.
[0057] A second example embodiment includes the features of the first example embodiment, wherein the plurality of routes for the plurality of vehicles of the optimized solutions are distributed across multiple days.
[0058] A third example embodiment includes the features of the first example embodiment, wherein the determining the compatible slot of the set comprises determining a plurality of compatible slots providing multiple time windows available to the order reservation.
[0059] A fourth example embodiment includes the features of the first example embodiment, wherein the calculating a set of permutations of slots comprises: identifying in the snapshot any vehicle plans that are present in each of a plurality of time windows, calculating a subset of the vehicle plans that are present as closest vehicle plans, calculating permutations for adding the order reservation to the set of closest vehicle plans, and identifying one of the permutations as a serviceable permutation.
[0060] A fifth example embodiment includes the features of the fourth example embodiment wherein the calculating a subset of the vehicle plans that are present as closest vehicle plans comprises evaluating a distance between a destination of the order reservation and one or more nodes of the vehicle plans that are present.
[0061] A sixth example embodiment includes the features of the fourth example embodiment, wherein the calculating permutations for adding the order reservation to the set of closest vehicle plans comprises calculating all possible permutations for adding the order reservation to the set of closest vehicle plans.
[0062] A seventh example embodiment includes the features of the first example embodiment, wherein the selecting for batching and dispatch is responsive to at least one of a state of solver and an update status of the optimized solutions.
[0063] An eighth example embodiment is a system comprising: a vehicle routing problem (VRP) optimization platform including a VRP solver configured to repeatedly operate to determine an optimized solutions of a non-deterministic polynomial-time hard vehicle routing problem, the optimized solutions defining a plurality of routes for a plurality of vehicles, each of the routes including a plurality of destinations for a respective vehicle of the plurality of vehicles and a respective deliverable for each of the plurality of destinations; and a real-time incorporation platform operatively coupled with the VRP optimization platform configured to: receive from an order processing platform, at a point during the repeatedly operating the solver, an order reservation from an order processing platform, the order reservation indicating at least one of an item and a service and including a new deliverable and destination, the new deliverable and destination not being included in any optimized solutions determined by the solver at said point,obtain via a optimization snapshot interface a snapshot of a current state of the optimized solutions of the solver, calculate a set of permutations of slots available in the snapshot for the new deliverable and destination, determine a compatible slot of the set, the compatible slot conforming to one or more optimization constraints applicable to the snapshot, provisionally add the new deliverable and destination to the compatible slot of the snapshot and to optimization parameters utilized by the solver, receive an order confirmation from the order processing platform, in response to the order confirmation, select for batching and dispatch one of (a) the snapshot with the new deliverable and destination added to the compatible slot and (b) a new optimized solutions determined by the solver with the new deliverable and destination included as optimization parameters, and batch a vehicle dispatch plan using the selected one of (a) and (b), and dispatching a plurality of vehicles according to the vehicle dispatch plan.
[0064] A ninth example embodiment includes the features of the eighth example embodiment, wherein the plurality of routes for the plurality of vehicles of the optimized solutions are distributed across multiple days.
[0065] A tenth example embodiment includes the features of the eighth example embodiment, wherein the real-time incorporation platform is configured to determine the compatible slot of the set by determining a plurality of compatible slots providing multiple time windows available to the order reservation.
[0066] An eleventh example embodiment includes the features of the eighth example embodiment, wherein the real-time incorporation platform is configured to calculate a set of permutations of slots by: identifying in the snapshot any vehicle plans that are present in each of a plurality of time windows, calculating a subset of the vehicle plans that are present as closest vehicle plans, calculating permutations for adding the order reservation to the set of closest vehicle plans, and identifying one of the permutations as a serviceable permutation.
[0067] A twelfth example embodiment includes the features of the eleventh example embodiment, wherein the calculating a subset of the vehicle plans that are present as closest vehicle plans comprises evaluating a distance between a destination of the order reservation and one or more nodes of the vehicle plans that are present.
[0068] A thirteenth example embodiment includes the features of the eleventh example embodiment, wherein the calculating permutations for adding the order reservation to the set ofclosest vehicle plans comprises calculating all possible permutations for adding the order reservation to the set of closest vehicle plans.
[0069] A fourteenth example embodiment includes the features of the eighth example embodiment, wherein the real-time incorporation platform is configured to perform a selection for batching and dispatch in response to at least one of a state of solver and an update status of the optimized solutions.
[0070] A fifteenth example embodiment is an apparatus comprising: at least one nontransitory memory medium configured to store instructions executable by one or more processors to perform the acts of: repeatedly operating a solver to determine an optimized solutions of a non- deterministic polynomial-time hard vehicle routing problem, the optimized solutions defining a plurality of routes for a plurality of vehicles, each of the routes including a plurality of destinations for a respective vehicle of the plurality of vehicles and a respective deliverable for each of the plurality of destinations; at a point during the repeatedly operating the solver, receiving an order reservation from an order processing platform, the order reservation indicating at least one of an item and a service and including a new deliverable and destination, the new deliverable and destination not being included in any optimized solutions determined by the solver at said point; obtaining a snapshot of a current state of the optimized solutions of the solver; calculating a set of permutations of slots available in the snapshot for the new deliverable and destination; determining a compatible slot of the set, the compatible slot conforming to one or more optimization constraints applicable to the snapshot; provisionally adding the new deliverable and destination to the compatible slot of the snapshot and to optimization parameters utilized by the solver; after receiving the order reservation, receiving an order confirmation from the order processing platform; and in response to the order confirmation, selecting for batching and dispatch one of (a) the snapshot with the new deliverable and destination added to the compatible slot and (b) a new optimized solutions determined by the solver with the new deliverable and destination included as optimization parameters; batching a vehicle dispatch plan using the selected one of (a) and (b); and dispatching a plurality of vehicles according to the vehicle dispatch plan.
[0071] A sixteenth example embodiment includes the features of the fifteenth example embodiment, wherein the plurality of routes for the plurality of vehicles of the optimized solutions are distributed across multiple days.
[0072] A seventeenth example embodiment includes the features of the fifteenth example embodiment, wherein the determining the compatible slot of the set comprises determining a plurality of compatible slots providing multiple time windows available to the order reservation.
[0073] An eighteenth example embodiment includes the features of the fifteenth example embodiment, wherein the calculating a set of permutations of slots comprises: identifying in the snapshot any vehicle plans that are present in each of a plurality of time windows, calculating a subset of the vehicle plans that are present as closest vehicle plans, calculating permutations for adding the order reservation to the set of closest vehicle plans, and identifying one of the permutations as a serviceable permutation.
[0074] A nineteenth example embodiment includes the features of the eighteenth example embodiment, wherein the calculating a subset of the vehicle plans that are present as closest vehicle plans comprises evaluating a distance between a destination of the order reservation and one or more nodes of the vehicle plans that are present.
[0075] A twentieth example embodiment includes the features of the eighteenth example embodiment, wherein the calculating permutations for adding the order reservation to the set of closest vehicle plans comprises calculating all possible permutations for adding the order reservation to the set of closest vehicle plans.
[0076] A twenty-first example embodiment includes the features of the fifteenth example embodiment, wherein the selecting for batching and dispatch is responsive to at least one of a state of solver and an update status of the optimized solutions.
[0077] It shall be appreciated that terms such as “a non-transitory memory,” “a non-transitory memory medium,” and “a non-transitory memory device” refer to a number of types of devices and storage mediums which may be configured to store information, such as data or instructions, readable or executable by a processor or other components of a computer system and that such terms include and encompass a single or unitary device or medium storing such information, multiple devices or media across or among which respective portions of such information are stored, and multiple devices or media across or among which multiple copies of such information are stored.
[0078] It shall be appreciated that terms such as “determine,” “determined,” “determining” and the like when utilized in connection with a control method or process, an electronic control system or controller, electronic controls, or components or operations of the foregoing referinclusively to a number of acts, configurations, devices, operations, and techniques including, without limitation, calculation or computation of a parameter or value, obtaining a parameter or value from a lookup table or using a lookup operation, receiving parameters or values from a datalink or network communication, receiving an electronic signal indicative of the parameter or value, receiving output of a sensor indicative of the parameter or value, receiving other outputs or inputs indicative of the parameter or value, reading the parameter or value from a memory location on a computer-readable medium, receiving the parameter or value as a run-time parameter, and / or by receiving a parameter or value by which the interpreted parameter can be calculated, and / or by referencing a default value that is interpreted to be the parameter value.
[0079] While example embodiments of the disclosure have been illustrated and described in detail in the drawings and foregoing description, the same is to be considered as illustrative and not restrictive in character, it being understood that only certain example embodiments have been shown and described and that all changes and modifications that come within the spirit of the claimed inventions are desired to be protected. It should be understood that while the use of words such as preferable, preferably, preferred or more preferred utilized in the description above indicates that the feature so described may be more desirable, it nonetheless may not be necessary and embodiments lacking the same may be contemplated as within the scope of the invention, the scope being defined by the claims that follow. In reading the claims, it is intended that when words such as “a,” “an,” “at least one,” or “at least one portion” are used there is no intention to limit the claim to only one item unless specifically stated to the contrary in the claim. When the language “at least a portion” and / or “a portion” is used the item can include a portion and / or the entire item unless specifically stated to the contrary.
Claims
CLAIMS1. A process comprising: repeatedly operating a solver to determine an optimized solutions of a non-deterministic polynomial-time hard vehicle routing problem, the optimized solutions defining a plurality of routes for a plurality of vehicles, each of the routes including a plurality of destinations for a respective vehicle of the plurality of vehicles and a respective deliverable for each of the plurality of destinations; at a point during the repeatedly operating the solver, receiving an order reservation from an order processing platform, the order reservation indicating at least one of an item and a service and including a new deliverable and destination, the new deliverable and destination not being included in any optimized solutions determined by the solver at said point; obtaining a snapshot of a current state of the optimized solutions of the solver; calculating a set of permutations of slots available in the snapshot for the new deliverable and destination; determining a compatible slot of the set, the compatible slot conforming to one or more optimization constraints applicable to the snapshot; provisionally adding the new deliverable and destination to the compatible slot of the snapshot and to optimization parameters utilized by the solver; after receiving the order reservation, receiving an order confirmation from the order processing platform; and in response to the order confirmation, selecting for batching and dispatch one of (a) the snapshot with the new deliverable and destination added to the compatible slot and (b) a new optimized solutions determined by the solver with the new deliverable and destination included as optimization parameters; batching a vehicle dispatch plan using the selected one of (a) and (b); and dispatching a plurality of vehicles according to the vehicle dispatch plan.
2. The process of claim 1, wherein the plurality of routes for the plurality of vehicles of the optimized solutions are distributed across multiple days.
3. The process of claim 1, wherein the determining the compatible slot of the set comprises determining a plurality of compatible slots providing multiple time windows available to the order reservation.
4. The process of claim 1, wherein the calculating a set of permutations of slots comprises: identifying in the snapshot any vehicle plans that are present in each of a plurality of time windows, calculating a subset of the vehicle plans that are present as closest vehicle plans, calculating permutations for adding the order reservation to the set of closest vehicle plans, and identifying one of the permutations as a serviceable permutation.
5. The process of claim 4, wherein the calculating a subset of the vehicle plans that are present as closest vehicle plans comprises evaluating a distance between a destination of the order reservation and one or more nodes of the vehicle plans that are present.
6. The process of claim 4, wherein the calculating permutations for adding the order reservation to the set of closest vehicle plans comprises calculating all possible permutations for adding the order reservation to the set of closest vehicle plans.
7. The process of claim 1, wherein the selecting for batching and dispatch is responsive to at least one of a state of solver and an update status of the optimized solutions.
8. A system comprising: a vehicle routing problem (VRP) optimization platform including a VRP solver configured to repeatedly operate to determine an optimized solutions of a non-deterministic polynomial-time hard vehicle routing problem, the optimized solutions defining a plurality of routes for a plurality of vehicles, each of the routes including a plurality of destinations for a respective vehicle of the plurality of vehicles and a respective deliverable for each of the plurality of destinations; and a real-time incorporation platform operatively coupled with the VRP optimization platform configured to:receive from an order processing platform, at a point during the repeatedly operating the solver, an order reservation from an order processing platform, the order reservation indicating at least one of an item and a service and including a new deliverable and destination, the new deliverable and destination not being included in any optimized solutions determined by the solver at said point, obtain via a optimization snapshot interface a snapshot of a current state of the optimized solutions of the solver, calculate a set of permutations of slots available in the snapshot for the new deliverable and destination, determine a compatible slot of the set, the compatible slot conforming to one or more optimization constraints applicable to the snapshot, provisionally add the new deliverable and destination to the compatible slot of the snapshot and to optimization parameters utilized by the solver, receive an order confirmation from the order processing platform, in response to the order confirmation, select for batching and dispatch one of (a) the snapshot with the new deliverable and destination added to the compatible slot and (b) a new optimized solutions determined by the solver with the new deliverable and destination included as optimization parameters, and batch a vehicle dispatch plan using the selected one of (a) and (b), and dispatching a plurality of vehicles according to the vehicle dispatch plan.
9. The system of claim 8, wherein the plurality of routes for the plurality of vehicles of the optimized solutions are distributed across multiple days.
10. The system of claim 8, wherein the real-time incorporation platform is configured to determine the compatible slot of the set by determining a plurality of compatible slots providing multiple time windows available to the order reservation.
11. The system of claim 8, wherein the real-time incorporation platform is configured to calculate a set of permutations of slots by:identifying in the snapshot any vehicle plans that are present in each of a plurality of time windows, calculating a subset of the vehicle plans that are present as closest vehicle plans, calculating permutations for adding the order reservation to the set of closest vehicle plans, and identifying one of the permutations as a serviceable permutation.
12. The system of claim 11, wherein the calculating a subset of the vehicle plans that are present as closest vehicle plans comprises evaluating a distance between a destination of the order reservation and one or more nodes of the vehicle plans that are present.
13. The system of claim 11, wherein the calculating permutations for adding the order reservation to the set of closest vehicle plans comprises calculating all possible permutations for adding the order reservation to the set of closest vehicle plans.
14. The system of claim 8, wherein the real-time incorporation platform is configured to perform a selection for batching and dispatch in response to at least one of a state of solver and an update status of the optimized solutions.
15. An apparatus comprising: at least one nontransitory memory medium configured to store instructions executable by one or more processors to perform the acts of: repeatedly operating a solver to determine an optimized solutions of a non-deterministic polynomial-time hard vehicle routing problem, the optimized solutions defining a plurality of routes for a plurality of vehicles, each of the routes including a plurality of destinations for a respective vehicle of the plurality of vehicles and a respective deliverable for each of the plurality of destinations; at a point during the repeatedly operating the solver, receiving an order reservation from an order processing platform, the order reservation indicating at least one of an item and a service and including a new deliverable and destination, the new deliverable and destination not being included in any optimized solutions determined by the solver at said point;obtaining a snapshot of a current state of the optimized solutions of the solver; calculating a set of permutations of slots available in the snapshot for the new deliverable and destination; determining a compatible slot of the set, the compatible slot conforming to one or more optimization constraints applicable to the snapshot; provisionally adding the new deliverable and destination to the compatible slot of the snapshot and to optimization parameters utilized by the solver; after receiving the order reservation, receiving an order confirmation from the order processing platform; and in response to the order confirmation, selecting for batching and dispatch one of (a) the snapshot with the new deliverable and destination added to the compatible slot and (b) a new optimized solutions determined by the solver with the new deliverable and destination included as optimization parameters; batching a vehicle dispatch plan using the selected one of (a) and (b); and dispatching a plurality of vehicles according to the vehicle dispatch plan.
16. The apparatus of claim 15, wherein the plurality of routes for the plurality of vehicles of the optimized solutions are distributed across multiple days.
17. The apparatus of claim 15, wherein the determining the compatible slot of the set comprises determining a plurality of compatible slots providing multiple time windows available to the order reservation.
18. The apparatus of claim 15, wherein the calculating a set of permutations of slots comprises: identifying in the snapshot any vehicle plans that are present in each of a plurality of time windows, calculating a subset of the vehicle plans that are present as closest vehicle plans, calculating permutations for adding the order reservation to the set of closest vehicle plans, and identifying one of the permutations as a serviceable permutation.
19. The apparatus of claim 18, wherein the calculating a subset of the vehicle plans that are present as closest vehicle plans comprises evaluating a distance between a destination of the order reservation and one or more nodes of the vehicle plans that are present.
20. The apparatus of claim 18, wherein the calculating permutations for adding the order reservation to the set of closest vehicle plans comprises calculating all possible permutations for adding the order reservation to the set of closest vehicle plans.
21. The apparatus of claim 15, wherein the selecting for batching and dispatch is responsive to at least one of a state of solver and an update status of the optimized solutions.
Citation Information
Patent Citations
Systems and methods for automated real-time and advisory routing within a fleet of geographically distributed drivers
US20180315319A1
System and method for routing optimization
US20200124429A1
Systems and methods for determining delivery time and route assignments
US20200143319A1
