Why Logistic and Shipping Companies Control Tower Needs Custom Routing Intelligence

Re-Optimization: NEMT Scheduling

To constantly manage the changing, real-world logistics operations, control towers need to re-optimize supply chain processes. Consider a Non-Emergency Medical Transportation (NEMT) provider managing daily appointments for 3 vehicles servicing a mix of dialysis, physical therapy, and outpatient visits.

Early in the day, an optimized route plan for 6 patients’ pickup and dropoff is generated and dispatched to drivers. According to the suggested route plan, only 2 vans were used out of the 3 available to complete all the patient drop-offs. However, shortly after dispatching this route, a few changes occur:

• Two patients cancel their trips after postponing appointments to different dates.
  
  
• A nearby clinic calls in with two new urgent pickups for same-day appointments.
  
  

Let’s understand how the Re-Optimization feature of Route Optimization API overcomes dynamic changes to the routing constraints. Use the re-optimization method along with the unique ID of the original request to accommodate the new changes – adding 2 urgent patient pickups & dropoff details and removing two shipments which were cancelled – into the existing route plan.

Interpreting the Re-Optimized Route Plan:

• The re-optimized route plan used all three vehicles to accommodate new urgent appointments.
  
  
• The changes to the original plan were intelligently accommodated, with NEMT Van 2 retained pickups 3, 4, and NEMT Van 1 assigned no pickups initially.
  
  
• Patients 1 & 2 cancelled their trips, and those stops were removed from the vehicle’s route plan.
  
  
• NEMT Van 1 was used to handle urgent pickups without adding more stops.
  
  
• The original route plan for NEMT Van 3 was retained as-is, as its pickups were not directly affected by the changing constraints.
  
  

Task Sequencing (Relations)

Containers can’t leave port until customs is cleared, sequencing rules are exactly what control towers enforce. Relations are used to specify the association between different tasks that should be performed in a certain order. NextBillion.ai’s Route Optimization Flexible API offers three types of relations that the users can enforce:

• in_same_route
  
  
• in_sequence
  
  
• in_direct_sequence
  
  

Jobs & Shipments:
Start by defining 10 jobs and 3 shipments. Let’s define the properties of these tasks:

• A unique identifier for each task
  
  
• Location indexes for each task
  
  
• Specify the schedule of tasks
  
  

Vehicles: Next, add 4 vehicles that will be responsible for fulfilling the tasks within the defined constraints. To describe the vehicles and their properties, add:

• A unique ID for each vehicle
  
  
• Vehicle shift time or the time window
  
  
• Capacity to denote the amount of load that the vehicle can take
  
  
• Start_index to denote the point from where the vehicle would start.
  
  
• Costs for all vehicles. We have specified fixed cost for all vehicles
  
  

Locations: Next, define the locations object and add all the locations used in the problem along with a valid id.

Relations: And, lastly, define the relations between the tasks of our choice. For this example:

• Keep shipment 3, jobs 7,8,9 in same route
  
  
• Keep shipment 2 and job 10, 5 in direct sequence
  
  
• Keep jobs 6, 4, 1 in sequence
  
  
• Jobs 2, 3 and shipment 1 are not part of any relation
  
  

The Solution:

• The in_sequence relationship with jobs 6, 4, 1 is fulfilled by vehicle 1. Job 4 & 1 were never fulfilled within their time window.
  
  
• The in_direct_sequence relationship with shipment 2 and jobs 10 & 5 is fulfilled by vehicle 2. Job 5 was never fulfilled within its time window.
  
  
• The in_same_route relationship with shipment 3 and jobs 7, 8, 9 is fulfilled by vehicle 3.
  
  
• The rest of the other jobs – jobs 2, 3 and shipment 1 – were also fulfilled by vehicle 3 on the same route.
  
  

Maximum Waiting Time

Avoid trucks waiting at borders or depots beyond thresholds to prevent costly detention and idle labor. NextBillion.ai’s Route Optimization Flexible API provides options to control how long the driver must wait after arriving at the task’s location. This feature is helpful when users want to manage the wait times of their drivers or when achieving the lowest wait times is a crucial business goal (for example, consumer-facing services).

Setup: Start by defining 5 jobs and 1 shipment. For these tasks we add:

• A unique identifier for each task
  
  
• Location indexes for each task
  
  
• Specify the schedule of tasks
  
  
• Skills for each task
  
  
• The actual time taken to complete the tasks once the driver/vehicle is at the task’s location i.e. the service time for each task
  
  

Vehicles: Next, add 2 vehicles that are going to fulfill the tasks within the defined constraints. To describe the vehicles and their properties, add:

• A unique ID for each vehicle
  
  
• Vehicle shift time or the time window
  
  
• Capacity to denote the amount of load that the vehicle can take
  
  
• Start_index to denote the point from where the vehicle would start. Alternatively, the vehicles could also be assigned to a depot
  
  
• Costs for all vehicles
  
  

Locations: Next, define the locations object and add all the locations used in the problem along with a valid id.

Options: Lastly, enable the max_activity_waiting_time constraint for the problem. Configure the constraint to keep the waiting time below 8 minutes for each activity.

The Solution:

• The overall result shows a total distance of 24,826 meters covered in 2,276 seconds of driving time to complete all assigned tasks.
  
  
• The total wait time for both vehicles was 718 seconds (about 12 minutes).
  
  
• Three tasks were left unassigned because their wait times exceeded the 8-minute limit.
  
  
• The maximum wait time strictly follows the set limit, which can sometimes result in some tasks not being assigned.
  
  

Transportation Cost Optimization

Control towers need to balance cost, speed, and service level, so APIs that optimize costs while respecting these constraints fit well. The NextBillion.ai Route Optimization API allows you to model real-world business rules. For each vehicle in the plan, you can set any combination of fixed, per hour, per kilometer, and per order costs. Once set, routes with lower costs are generally preferred, while still considering other factors like time windows and task locations.

Setup Example:

• Locations: Add the locations used in the problem along with a valid id.
  
  
• Jobs & Shipments: Define 3 pickup jobs and 2 shipments. For each of these tasks, configure:
  
  
  • A unique identifier
    
    
  • A Location index
    
    
  • Specify the schedule of tasks using the time_windows
    
    
  • A service time
    
    

Vehicles: Add the 3 vans that are going to fulfill the tasks along with their cost components. Configure each vehicle with:

• • • A unique ID
      
      
    • A shift time or the time window
      
      
    • Capacity to denote the amount of load that the van can take
      
      
    • Start_index and end_indexes to denote the start and end points for a van
      
      
    • Skills of the driver or the van
      
      

Cost Types:

• Van 1: A per_hour cost
  
  
• Van 2: A per_order cost
  
  
• Van 3: A fixed cost
  
  

The Solution:

• Vehicle 1: Costs $505 for a half-hour route at 1000 per hour.
  
  
• Vehicle 2: Costs $3,515 for a single order plus travel time.
  
  
• Vehicle 3: Costs $5,027 fixed regardless of distance.