Cost Optimization

The VRP solver provides comprehensive cost modeling to optimize your routing operations based on real business constraints. This guide covers the recommended cost-based configuration, weight-based tuning, and optimization strategies.

Quick Start: Cost-Based Optimization (Recommended)

The simplest way to configure the solver is using the costs configuration. Instead of tuning abstract weights, you specify actual business costs in EUR (or your currency), and the solver derives optimal weights automatically.

{
  "resources": [
    {
      "name": "driver-1",
      "hourlyCost": 25,
      "shifts": [{
        "from": "2024-03-15T08:00:00Z",
        "to": "2024-03-15T17:00:00Z"
      }]
    }
  ],
  "jobs": [
    {
      "name": "delivery-1",
      "location": {"latitude": 52.520, "longitude": 13.405},
      "duration": 1800
    }
  ],
  "costs": {
    "drivingCostPerHour": 25.0,
    "distanceCostPerKm": 0.35,
    "waitingCostPerHour": 15.0
  }
}

Why use costs instead of weights?

Intuitive: Costs represent real business values (EUR/hour, EUR/km)
Comparable: Different constraint types can be directly compared
Transparent: Response includes a cost breakdown in your currency
Maintainable: Business changes are easier to translate to configuration

Available Cost Parameters

Parameter	Description	Default	Unit
`drivingCostPerHour`	Cost of driver time while driving/traveling	25.0	EUR/hour
`distanceCostPerKm`	Cost per kilometer (fuel, maintenance, wear)	0.0	EUR/km
`waitingCostPerHour`	Cost of idle/waiting time	0.0	EUR/hour
`overtimeCostPerHour`	Cost of overtime work	50.0	EUR/hour
`timeWindowViolationCostPerHour`	Penalty for missing time windows	20.0	EUR/hour
`priorityCostPerPointPerHour`	Cost per priority point per hour late	5.0	EUR
`preferredResourceViolationCost`	Penalty for not using preferred resource	20.0	EUR
`rankingViolationCostPerRank`	Cost per rank deviation	5.0	EUR
`resourceActivationCost`	Fixed cost applied once per activated resource. This is a single request-wide value — there is no per-resource activation cost field	0.0 (disabled)	EUR

Cost Breakdown in Response

When using costs, the response includes a detailed financial breakdown:

{
  "score": {"hardScore": 0, "mediumScore": 0, "softScore": -125000, "feasible": true},
  "trips": [...],
  "totalTravelTimeInSeconds": 7200,
  "totalTravelDistanceInMeters": 50000,
  "estimatedCost": {
    "totalCostEur": 85.25,
    "travelTimeCostEur": 50.00,
    "distanceCostEur": 17.50,
    "waitTimeCostEur": 3.75,
    "laborCostEur": 12.50,
    "activationCostEur": 0.00,
    "penaltyCostEur": 1.50
  }
}

Cost Types Overview

The solver considers multiple cost components when optimizing routes:

Total Cost = Fixed Costs + Variable Costs + Penalty Costs

Fixed Costs: Activation costs for using resources
Variable Costs: Time-based costs (hourly rates, overtime), distance costs
Penalty Costs: Violations of soft constraints

Resource Cost Configuration

Hourly Costs

Configure time-based costs for resources:

{
  "resources": [
    {
      "name": "employee-driver",
      "hourlyCost": 25,  // $25/hour
      "shifts": [{
        "from": "2024-03-15T08:00:00Z",
        "to": "2024-03-15T17:00:00Z"
      }]
    },
    {
      "name": "contractor-driver",
      "hourlyCost": 45,  // $45/hour
      "shifts": [{
        "from": "2024-03-15T08:00:00Z",
        "to": "2024-03-15T17:00:00Z"
      }]
    }
  ],
  "jobs": [
    {
      "name": "delivery-1",
      "duration": 1800
    }
  ]
}

Hourly costs apply to the entire shift duration, not just driving time. Consider using activation costs if you only want to charge for actual usage.

Activation Costs

A fixed cost incurred whenever a resource is actually used. This is configured once for the whole request via costs.resourceActivationCost — there is no per-resource activationCost field on the resource itself:

{
  "resources": [
    {
      "name": "rental-truck",
      "hourlyCost": 0,      // No hourly charge
      "shifts": [{
        "from": "2024-03-15T08:00:00Z",
        "to": "2024-03-15T17:00:00Z"
      }]
    },
    {
      "name": "owned-vehicle",
      "hourlyCost": 15,      // $15/hour driver cost
      "shifts": [{
        "from": "2024-03-15T08:00:00Z",
        "to": "2024-03-15T17:00:00Z"
      }]
    }
  ],
  "jobs": [
    {
      "name": "job-1",
      "duration": 1800
    }
  ],
  "costs": {
    "resourceActivationCost": 150  // Applied once per resource that ends up being used
  }
}

Cost Calculation Examples

Because resourceActivationCost is a single request-wide value, it applies identically to every resource that ends up used — the only per-resource lever for cost differences is hourlyCost:

Employee Model
Contractor Model
Rental Model

{
  "resources": [{"name": "full-time-driver", "hourlyCost": 30}],
  "costs": {"resourceActivationCost": 0}
}

8-hour shift cost: 8 ×

30 =

240 Used or not: Same cost (salaried, no activation cost)

{
  "resources": [{"name": "gig-driver", "hourlyCost": 40}],
  "costs": {"resourceActivationCost": 25}
}

5-hour usage:

25 + (5 ×

40) =

225 **Incentive**: Use fewer contractors, since every one activated adds the flat

{
  "resources": [{"name": "rental-van", "hourlyCost": 0}],
  "costs": {"resourceActivationCost": 200}
}

Any usage: Flat $200 Strategy: Maximize utilization once activated

resourceActivationCost lives under costs and applies request-wide — you cannot give individual resources different activation costs.

Travel Cost Optimization

Using Cost-Based Configuration (Recommended)

The easiest way to optimize travel costs is using the costs configuration:

{
  "costs": {
    "drivingCostPerHour": 25.0,
    "distanceCostPerKm": 0.35
  }
}

This example says:

Every hour of driving costs 25 EUR (driver wages, opportunity cost)
Every kilometer driven costs 0.35 EUR (fuel, maintenance, depreciation)

The solver will balance both time and distance to minimize total cost.

Typical cost values:

Driving cost: 20-40 EUR/hour (depending on driver wages)
Distance cost: 0.25-0.50 EUR/km (depending on vehicle type)
Small city car: ~0.25 EUR/km
Delivery van: ~0.35 EUR/km
Heavy truck: ~0.50 EUR/km

Using Weight-Based Configuration (Advanced)

For fine-grained control, use the weights configuration:

{
  "weights": {
    "travelTimeWeight": 1,
    "waitTimeWeight": 0
  }
}

number

default:"1"

Multiplier for travel time in the objective function. Higher values prioritize shorter travel times.

Balancing Travel vs Other Objectives

{
  "weights": {
    "travelTimeWeight": 10.0,
    "waitTimeWeight": 1.0,
    "urgencyWeight": 1.0
  }
}

{
  "weights": {
    "travelTimeWeight": 1.0,
    "waitTimeWeight": 1.0,
    "urgencyWeight": 5.0
  }
}

{
  "weights": {
    "travelTimeWeight": 0.5,
    "waitTimeWeight": 0.5,
    "urgencyWeight": 20.0
  }
}

Urgency-Based Optimization

Basic Urgency Scoring

Prioritize time-sensitive jobs:

{
  "jobs": [
    {
      "name": "critical-repair",
      "urgency": 100,  // Highest priority
      "location": {"latitude": 52.520, "longitude": 13.405},
      "duration": 3600
    },
    {
      "name": "routine-maintenance",
      "urgency": 10,   // Lower priority
      "location": {"latitude": 52.520, "longitude": 13.405},
      "duration": 1800
    }
  ],
  "resources": [
    {
      "name": "technician-1",
      "shifts": [{
        "from": "2024-03-15T08:00:00Z",
        "to": "2024-03-15T17:00:00Z"
      }]
    }
  ],
  "weights": {
    "urgencyWeight": 50  // Strong urgency preference
  }
}

integer

default:"0"

Priority score from 0-100. Higher values are scheduled earlier.

Urgency Calculation Formula

Urgency Impact = urgency × urgencyWeight × time_until_scheduledThis creates increasing pressure to schedule urgent jobs sooner.

ASAP Scheduling

Schedule jobs as soon as possible across multiple days:

{
  "jobs": [
    {
      "name": "flexible-task-1",
      "windows": [
        {"from": "2024-03-15T08:00:00Z", "to": "2024-03-15T17:00:00Z"},
        {"from": "2024-03-16T08:00:00Z", "to": "2024-03-16T17:00:00Z"},
        {"from": "2024-03-17T08:00:00Z", "to": "2024-03-17T17:00:00Z"}
      ]
    },
    {
      "name": "flexible-task-2",
      "windows": [
        {"from": "2024-03-16T08:00:00Z", "to": "2024-03-16T17:00:00Z"},
        {"from": "2024-03-17T08:00:00Z", "to": "2024-03-17T17:00:00Z"}
      ]
    }
  ],
  "resources": [
    {
      "name": "driver-1",
      "shifts": [
        {
          "from": "2024-03-15T08:00:00Z",
          "to": "2024-03-15T17:00:00Z"
        },
        {
          "from": "2024-03-16T08:00:00Z",
          "to": "2024-03-16T17:00:00Z"
        },
        {
          "from": "2024-03-17T08:00:00Z",
          "to": "2024-03-17T17:00:00Z"
        }
      ]
    }
  ],
  "weights": {
    "asapWeight": 100  // Strong ASAP preference
  }
}

ASAP Strategy

There is no dayIndex job field. The earliest day a job can be scheduled is determined entirely by its windows: flexible-task-1’s first window opens on day 1, while flexible-task-2 only becomes available on day 2 because that’s where its first window starts.

Express Earliest Availability via Windows

A job’s first windows entry sets its earliest possible day — omit early windows to push earliest availability out

Configure Weight

Higher asapWeight = stronger preference for using a job’s earliest available window over a later one

Combine with Urgency

{
  "urgency": 80,
  "windows": [
    {"from": "2024-03-15T08:00:00Z", "to": "2024-03-15T17:00:00Z"}
  ]
}

Optimization Strategies

Strategy 1: Cost Minimization

Focus on reducing operational costs:

{
  "resources": [
    {
      "name": "efficient-vehicle",
      "hourlyCost": 20,
      "capacity": [1000],
      "shifts": [{
        "from": "2024-03-15T08:00:00Z",
        "to": "2024-03-15T17:00:00Z"
      }]
    },
    {
      "name": "large-vehicle",
      "hourlyCost": 35,
      "capacity": [2000],
      "shifts": [{
        "from": "2024-03-15T08:00:00Z",
        "to": "2024-03-15T17:00:00Z"
      }]
    }
  ],
  "jobs": [
    {
      "name": "delivery-1",
      "load": [800]
    }
  ],
  "weights": {
    "travelTimeWeight": 0.5,       // Less important
    "minimizeResourcesWeight": 2.0 // Minimize vehicle use
  }
}

There is no per-resource activationCost field. A flat cost per activated resource is configured request-wide via costs.resourceActivationCost; the weight-based equivalent for reducing the number of vehicles used is weights.minimizeResourcesWeight.

Strategy 2: Service Level Optimization

Prioritize customer satisfaction:

{
  "resources": [
    {
      "name": "technician-1",
      "shifts": [{
        "from": "2024-03-15T08:00:00Z",
        "to": "2024-03-15T17:00:00Z"
      }]
    }
  ],
  "jobs": [
    {
      "name": "service-1",
      "urgency": 90,
      "duration": 3600,
      "windows": [
        {"from": "2024-03-15T09:00:00Z", "to": "2024-03-15T11:00:00Z", "hard": false, "weight": 50}
      ]
    }
  ],
  "weights": {
    "urgencyWeight": 100,          // Honor priorities
    "waitTimeWeight": 20,          // Minimize customer waiting
    "preferredVehicleWeight": 30   // Use preferred technicians
  }
}

There is no global windowWeight. Time-window preference strength is set per window via that window’s own hard/weight fields (hard: false makes it a soft preference, weight sets the penalty for missing it). The weight for keeping a job on its preferred/planned resource is preferredVehicleWeight, not preferredResourceWeight.

Strategy 3: Balanced Optimization

{
  "resources": [
    {
      "name": "driver-1",
      "shifts": [{
        "from": "2024-03-15T08:00:00Z",
        "to": "2024-03-15T17:00:00Z"
      }]
    }
  ],
  "jobs": [
    {
      "name": "job-1",
      "duration": 1800
    }
  ],
  "options": {
    // Fairness factors
    "fairWorkloadPerResource": true,
    "fairWorkloadPerTrip": true
  },
  "weights": {
    // Cost factors
    "travelTimeWeight": 1.0,
    "minimizeResourcesWeight": 1.0,
    "overtimeWeight": 50,

    // Service factors
    "urgencyWeight": 10,

    // Fairness factor (controls fairWorkloadPerTrip/fairWorkloadPerResource strength)
    "workloadSpreadWeight": 5
  }
}

fairWorkloadPerResource and fairWorkloadPerTrip are booleans under options, not numeric weights. The single weight that controls how aggressively both are enforced is weights.workloadSpreadWeight.

Cost Analysis in Results

When using the costs configuration, the response includes a detailed estimatedCost breakdown:

{
  "score": {"hardScore": 0, "mediumScore": 0, "softScore": -125000, "feasible": true},
  "trips": [...],
  "totalTravelTimeInSeconds": 14400,
  "totalTravelDistanceInMeters": 120000,
  "estimatedCost": {
    "totalCostEur": 245.75,
    "travelTimeCostEur": 100.00,
    "distanceCostEur": 42.00,
    "waitTimeCostEur": 8.25,
    "laborCostEur": 75.00,
    "activationCostEur": 20.00,
    "penaltyCostEur": 0.50
  }
}

The estimatedCost field is only included when you provide a costs configuration in the request.

Understanding Cost Components

Fixed Costs

Resource activation costs
Daily vehicle fees
Equipment rental charges
Base operating costs

Variable Costs

Hourly labor costs
Fuel costs (via travel time)
Overtime premiums
Distance-based charges

Penalty Costs

Soft time window violations
Preference violations
Workload imbalance penalties
Capacity overages (if allowed)

Advanced Cost Scenarios

Peak Hour Pricing

Different costs for different times:

{
  "resources": [
    {
      "name": "peak-driver",
      "shifts": [
        {
          "from": "2024-03-15T06:00:00Z",
          "to": "2024-03-15T10:00:00Z",
          "hourlyCost": 35  // Morning peak rate
        },
        {
          "from": "2024-03-15T10:00:00Z",
          "to": "2024-03-15T15:00:00Z",
          "hourlyCost": 25  // Standard rate
        },
        {
          "from": "2024-03-15T15:00:00Z",
          "to": "2024-03-15T19:00:00Z",
          "hourlyCost": 35  // Evening peak rate
        }
      ]
    }
  ],
  "jobs": [
    {
      "name": "delivery-1", 
      "duration": 1800
    }
  ]
}

Performance Tips

Cost Calculation Performance:

Complex cost models increase computation time
Many soft constraints require more iterations
Large weight differences can cause instability

Recommendations:

Keep weights in reasonable ratios (1:100 max)
Use hard constraints where possible
Profile different weight combinations

Best Practices

Start with Defaults

Begin with default weights and adjust based on results

Measure Impact

Track these metrics:

Total operational cost
Cost per delivery/service
Resource utilization rates
Customer satisfaction scores

Iterate and Refine

Adjust weights based on business priorities
Monitor unintended consequences
Document weight configurations

Consider Trade-offs

Lower costs vs better service
Fewer vehicles vs more overtime
Urgency vs efficiency

Troubleshooting

Unexpected High Costs

Check:

Overtime penalties being triggered
Excessive activation of resources
Long travel times due to poor clustering
Soft constraint violations

Solutions:

Review shift definitions
Adjust activation costs
Modify territory assignments
Convert critical soft constraints to hard

Poor Route Quality

Check:

Weight balance between objectives
Conflicting constraints
Insufficient resources

Solutions:

Reduce focus on cost minimization
Increase service-related weights
Add resources or extend shifts

Urgency Not Respected

Check:

urgencyWeight value
Competing objectives
Time window conflicts

Solutions:

Increase urgencyWeight significantly
Reduce other weight values
Review job time windows

Resource Management

Configure resource costs and constraints

Time Windows

Balance costs with time constraints

Workload Balancing

Fair distribution vs cost optimization

Scoring System

How costs affect solution scoring

Platform

VRP API

Fill API

Cost Optimization

Cost Optimization

Quick Start: Cost-Based Optimization (Recommended)

Available Cost Parameters

Cost Breakdown in Response

Cost Types Overview

Resource Cost Configuration

Hourly Costs

Activation Costs

Cost Calculation Examples

Travel Cost Optimization

Using Cost-Based Configuration (Recommended)

Using Weight-Based Configuration (Advanced)

Balancing Travel vs Other Objectives

Urgency-Based Optimization

Basic Urgency Scoring

Urgency Calculation Formula

ASAP Scheduling

ASAP Strategy

Optimization Strategies

Strategy 1: Cost Minimization

Strategy 2: Service Level Optimization

Strategy 3: Balanced Optimization

Cost Analysis in Results

Understanding Cost Components

Advanced Cost Scenarios

Peak Hour Pricing

Performance Tips

Best Practices

Troubleshooting

Resource Management

Time Windows

Workload Balancing

Scoring System

​Cost Optimization

​Quick Start: Cost-Based Optimization (Recommended)

​Available Cost Parameters

​Cost Breakdown in Response

​Cost Types Overview

​Resource Cost Configuration

​Hourly Costs

​Activation Costs

​Cost Calculation Examples

​Travel Cost Optimization

​Using Cost-Based Configuration (Recommended)

​Using Weight-Based Configuration (Advanced)

​Balancing Travel vs Other Objectives

​Urgency-Based Optimization

​Basic Urgency Scoring

​Urgency Calculation Formula

​ASAP Scheduling

​ASAP Strategy

​Optimization Strategies

​Strategy 1: Cost Minimization

​Strategy 2: Service Level Optimization

​Strategy 3: Balanced Optimization

​Cost Analysis in Results

​Understanding Cost Components

​Advanced Cost Scenarios

​Peak Hour Pricing

​Performance Tips

​Best Practices

​Troubleshooting

​Related Features

Resource Management

Time Windows

Workload Balancing

Scoring System

Cost Optimization

Quick Start: Cost-Based Optimization (Recommended)

Available Cost Parameters

Cost Breakdown in Response

Cost Types Overview

Resource Cost Configuration

Hourly Costs

Activation Costs

Cost Calculation Examples

Travel Cost Optimization

Using Cost-Based Configuration (Recommended)

Using Weight-Based Configuration (Advanced)

Balancing Travel vs Other Objectives

Urgency-Based Optimization

Basic Urgency Scoring

Urgency Calculation Formula

ASAP Scheduling

ASAP Strategy

Optimization Strategies

Strategy 1: Cost Minimization

Strategy 2: Service Level Optimization

Strategy 3: Balanced Optimization

Cost Analysis in Results

Understanding Cost Components

Advanced Cost Scenarios

Peak Hour Pricing

Performance Tips

Best Practices

Troubleshooting

Related Features