Job Priority

Overview

The priority field on jobs allows you to influence which jobs are more likely to be included in the final schedule when not all jobs can be served. This is particularly useful in overconstrained scenarios where you have more work than available resources.

Priority affects whether a job gets scheduled, not when it gets scheduled. Use time windows to control scheduling order.

How Priority Works

When the VRP solver evaluates solutions, unassigned jobs generate penalty scores. The penalty for each unserved job is calculated as:

penalty = priority × duration × priorityWeight

Higher penalties make the solver work harder to include those jobs in the schedule, effectively prioritizing them over jobs with lower penalties.

Key Concepts

Default Priority

If not specified, jobs default to priority 1.

Priority Multiplication

The priority value is multiplied by the job’s duration, meaning longer high-priority jobs have an even stronger influence on scheduling decisions.

Weight Configuration

The global priorityWeight parameter in the request weights controls how strongly priority influences the solution.

Basic Example

{
  "jobs": [
    {
      "name": "critical-delivery",
      "duration": 600,
      "priority": 100,
      "location": {
        "latitude": 51.0543,
        "longitude": 3.7174
      }
    },
    {
      "name": "standard-delivery",
      "duration": 600,
      "priority": 1,
      "location": {
        "latitude": 51.0456,
        "longitude": 3.7326
      }
    },
    {
      "name": "low-priority",
      "duration": 600,
      "priority": 0,
      "location": {
        "latitude": 51.0389,
        "longitude": 3.7203
      }
    }
  ],
  "resources": [
    {
      "name": "vehicle-1",
      "shifts": [
        {
          "from": "2023-01-13T09:00:00.000Z",
          "to": "2023-01-13T17:00:00.000Z",
          "start": {
            "latitude": 51.0500,
            "longitude": 3.7200
          }
        }
      ]
    }
  ],
  "options": {
    "partialPlanning": true
  },
  "weights": {
    "priorityWeight": 100
  }
}

Complete Job Example

Here’s a complete job example showing all available fields:

{
  "jobs": [
    {
      "name": "critical-medical-delivery",
      "priority": 100,
      "urgency": 80,
      "duration": 1800,
      "location": {
        "latitude": 51.0543,
        "longitude": 3.7174
      },
      "tags": [
        {
          "name": "medical",
          "hard": true
        },
        {
          "name": "fragile",
          "hard": false,
          "weight": 100
        }
      ],
      "windows": [
        {
          "from": "2023-01-13T09:00:00.000Z",
          "to": "2023-01-13T12:00:00.000Z"
        },
        {
          "from": "2023-01-13T14:00:00.000Z",
          "to": "2023-01-13T17:00:00.000Z"
        }
      ],
      "rankings": [
        {
          "resource": "driver-001",
          "value": 1
        }
      ],
      "disallowedResources": ["driver-trainee"],
      "hard": true,
      "hardWeight": 1000,
      "padding": 300,
      "load": [10, 5],
      "resumable": false
    }
  ]
}

The name field is the only required field for a job. All other fields, including priority, are optional.

Advanced Usage

Combining with Partial Planning

Priority works seamlessly with partial planning to handle overconstrained problems:

{
  "jobs": [
    {
      "name": "emergency-repair",
      "duration": 7200,
      "priority": 100,
      "location": {
        "latitude": 51.05,
        "longitude": 3.72
      }
    },
    {
      "name": "scheduled-maintenance-1",
      "duration": 3600,
      "priority": 20,
      "location": {
        "latitude": 51.06,
        "longitude": 3.73
      }
    },
    {
      "name": "scheduled-maintenance-2",
      "duration": 3600,
      "priority": 20,
      "location": {
        "latitude": 51.04,
        "longitude": 3.71
      }
    },
    {
      "name": "routine-check",
      "duration": 1800,
      "priority": 1,
      "location": {
        "latitude": 51.03,
        "longitude": 3.74
      }
    }
  ],
  "resources": [
    {
      "name": "technician-1",
      "shifts": [
        {
          "from": "2023-01-13T08:00:00.000Z",
          "to": "2023-01-13T16:00:00.000Z",
          "start": {
            "latitude": 51.05,
            "longitude": 3.72
          },
          "end": {
            "latitude": 51.05,
            "longitude": 3.72
          }
        }
      ]
    }
  ],
  "options": {
    "partialPlanning": true,
    "explain": true
  }
}

Priority Strategies

Emergency Response

{
  "jobs": [
    {
      "name": "emergency-1",
      "priority": 1000,
      "duration": 1800,
      "location": {
        "latitude": 51.05,
        "longitude": 3.72
      }
    },
    {
      "name": "urgent-1",
      "priority": 100,
      "duration": 2400,
      "location": {
        "latitude": 51.06,
        "longitude": 3.73
      }
    },
    {
      "name": "routine-1",
      "priority": 1,
      "duration": 3600,
      "location": {
        "latitude": 51.04,
        "longitude": 3.71
      }
    }
  ]
}

Configuration Guidelines

Priority Weight Tuning

The priorityWeight parameter controls how strongly priority influences the solution:

Low Weight (1-10)

Priority has minimal impact. The solver focuses more on efficiency metrics like travel time and workload balance.

Medium Weight (10-100)

Priority significantly influences job selection while still considering other optimization goals.

High Weight (100-1000)

Priority becomes dominant. High-priority jobs will almost always be scheduled before lower-priority ones.

Best Practices

Setting all jobs to the same high priority defeats the purpose. Use priority to create meaningful differentiation between job importance levels.

For jobs that must be completed, consider using hard constraints (like time windows) instead of relying solely on priority.

Understanding the Results

When using the explain option, you can see how priority affected unassigned jobs:

{
  "solution": {
    "routes": [
      {
        "resourceId": "technician-1",
        "shiftId": "shift1",
        "activities": [
          {
            "type": "start",
            "location": {"lat": 51.05, "lon": 3.72},
            "time": {
              "arrival": "2023-01-13T08:00:00Z",
              "departure": "2023-01-13T08:00:00Z"
            }
          },
          {
            "type": "job",
            "jobId": "emergency-repair",
            "location": {"lat": 51.05, "lon": 3.72},
            "time": {
              "arrival": "2023-01-13T08:00:00Z",
              "departure": "2023-01-13T10:00:00Z"
            }
          },
          {
            "type": "job",
            "jobId": "scheduled-maintenance-1",
            "location": {"lat": 51.06, "lon": 3.73},
            "time": {
              "arrival": "2023-01-13T10:30:00Z",
              "departure": "2023-01-13T11:30:00Z"
            }
          },
          {
            "type": "end",
            "location": {"lat": 51.05, "lon": 3.72},
            "time": {
              "arrival": "2023-01-13T12:00:00Z"
            }
          }
        ]
      }
    ],
    "unassigned": [
      {
        "jobId": "routine-check",
        "reasons": [
          {
            "code": "UNSERVED_JOBS",
            "description": "Job priority: 1, Duration: 1800s, Score: -180000"
          }
        ]
      }
    ]
  }
}

Common Use Cases

1. Service Level Agreements

Ensure SLA-critical jobs are prioritized:

{
  "resources": [
    {
      "id": "tech-1",
      "name": "Technician 1",
      "shifts": [{"id": "shift1", "from": "2023-01-13T08:00:00Z", "to": "2023-01-13T18:00:00Z"}]
    }
  ],
  "jobs": [
    {
      "id": "sla-critical-job",
      "name": "SLA Critical Job",
      "priority": 100,
      "serviceDurationInSeconds": 3600,
      "location": {"lat": 51.05, "lon": 3.72}
    },
    {
      "id": "standard-job",
      "name": "Standard Job",
      "priority": 10,
      "serviceDurationInSeconds": 3600,
      "location": {"lat": 51.06, "lon": 3.73}
    },
    {
      "id": "best-effort-job",
      "name": "Best Effort Job",
      "priority": 1,
      "serviceDurationInSeconds": 3600,
      "location": {"lat": 51.04, "lon": 3.71}
    }
  ]
}

2. Emergency Services

Handle emergency requests alongside scheduled work:

{
  "resources": [
    {
      "id": "tech-1",
      "name": "Emergency Technician",
      "shifts": [{"id": "shift1", "from": "2023-01-13T08:00:00Z", "to": "2023-01-13T18:00:00Z"}]
    }
  ],
  "jobs": [
    {
      "id": "emergency-response",
      "name": "Emergency Response",
      "priority": 1000,
      "serviceDurationInSeconds": 1800,
      "urgency": 100,
      "location": {"lat": 51.05, "lon": 3.72}
    },
    {
      "id": "urgent-repair",
      "name": "Urgent Repair",
      "priority": 50,
      "serviceDurationInSeconds": 3600,
      "urgency": 50,
      "location": {"lat": 51.06, "lon": 3.73}
    },
    {
      "id": "scheduled-maintenance",
      "name": "Scheduled Maintenance",
      "priority": 1,
      "serviceDurationInSeconds": 3600,
      "location": {"lat": 51.04, "lon": 3.71}
    }
  ]
}

3. Maximize Revenue

Prioritize high-value opportunities:

{
  "resources": [
    {
      "id": "installer-1",
      "name": "Installer 1",
      "shifts": [{"id": "shift1", "from": "2023-01-13T08:00:00Z", "to": "2023-01-13T18:00:00Z"}]
    }
  ],
  "jobs": [
    {
      "id": "high-value-installation",
      "name": "High Value Installation",
      "priority": 80,
      "serviceDurationInSeconds": 7200,
      "location": {"lat": 51.05, "lon": 3.72},
      "tags": [{"name": "revenue-8000", "hard": false}]
    },
    {
      "id": "medium-value-service",
      "name": "Medium Value Service",
      "priority": 35,
      "serviceDurationInSeconds": 3600,
      "location": {"lat": 51.06, "lon": 3.73},
      "tags": [{"name": "revenue-3500", "hard": false}]
    },
    {
      "id": "standard-service",
      "name": "Standard Service",
      "priority": 10,
      "serviceDurationInSeconds": 1800,
      "location": {"lat": 51.04, "lon": 3.71},
      "tags": [{"name": "revenue-1000", "hard": false}]
    }
  ]
}

Partial Planning

Allow solver to leave some jobs unassigned in overconstrained scenarios

Time Windows

Control when jobs can be scheduled

Job Urgency

Prioritize jobs to be scheduled earlier in the day

Constraints

Understand all available constraints

Platform

VRP API

Overview

How Priority Works

Key Concepts

Basic Example

Complete Job Example

Advanced Usage

Combining with Partial Planning

Priority Strategies

Configuration Guidelines

Priority Weight Tuning

Best Practices

Understanding the Results

Common Use Cases

1. Service Level Agreements

2. Emergency Services

3. Maximize Revenue

Partial Planning

Time Windows

Job Urgency

Constraints

Platform

VRP API

​Overview

​How Priority Works

​Key Concepts

​Basic Example

​Complete Job Example

​Advanced Usage

​Combining with Partial Planning

​Priority Strategies

​Configuration Guidelines

​Priority Weight Tuning

​Best Practices

​Understanding the Results

​Common Use Cases

​1. Service Level Agreements

​2. Emergency Services

​3. Maximize Revenue

​Related Features

Partial Planning

Time Windows

Job Urgency

Constraints

Overview

How Priority Works

Key Concepts

Basic Example

Complete Job Example

Advanced Usage

Combining with Partial Planning

Priority Strategies

Configuration Guidelines

Priority Weight Tuning

Best Practices

Understanding the Results

Common Use Cases

1. Service Level Agreements

2. Emergency Services

3. Maximize Revenue

Related Features