Terminology

This section lists frequently used terms and their meanings:

Scheduled periods within a driver shift when a driver is not available for jobs. Floating breaks (such as lunch breaks) can be taken at any location between stops. Fixed appointments occur at a specific time and location, with travel time included in the route. Learn more about lunch breaks and personal appointments.

The total load that can be transported by a driver’s vehicle at any one time. A vehicle can have more than one capacity type, for example, seats for ambulatory passengers and positions for wheelchairs. Learn more about driver capacity.

The hard, medium, and soft rules that must be taken into account when creating a pick-up and delivery schedule. Learn more about constraints.

In passenger transport, assigning a client to the same driver, or a small set of familiar drivers, over time. Continuity of care is modeled by setting preferred or required drivers on a client’s jobs. Learn more about preferred drivers.

Timefold accepts input datasets that include information about the drivers and the jobs to be scheduled and optimized. Input datasets are created using our predefined model schema which follows the OpenAPI specification. Timefold returns output datasets that contain the optimized solution. See the API spec for more information.

Distance a driver travels with an empty vehicle, for example, between a drop-off and the next pick-up. Also called empty miles. Route optimization and job pooling both reduce deadhead miles. Learn more about route optimization.

The amount of vehicle capacity consumed by a job’s stops at any given point in a route. When a pick-up stop is completed, the demand increases; when the corresponding delivery stop is completed, the demand is released. Learn more about driver capacity.

The rules that keep a job’s stops in the correct order and with the same driver. For example, an item can’t be delivered before it’s picked up, and the same driver must complete both stops. Learn more about dependencies between stops.

A person who can be assigned jobs. Drivers work shifts with defined start and end times, and can have skills and tags that determine which jobs they can be assigned.

A break that occurs at a specific start and end time. When a location is provided, Timefold includes travel time to and from the break in the route. When no location is provided, the time is blocked but travel time cannot be accounted for. Fixed breaks model personal appointments and other scheduled commitments. Learn more about fixed breaks.

A break with a defined duration that can be taken at any location within a specified time window. Timefold schedules floating breaks in around stops to minimize disruption to the route. Lunch breaks are a typical example. Learn more about floating breaks.

A series of stops that typically include a pick-up location and a delivery location, and can include additional service stops in between. All stops in a job must be completed by the same driver and in the order they are listed.

Completing stops from multiple jobs concurrently by interleaving them in a single route. For example, picking up two parcels before delivering both saves travel time compared to completing each job separately. In passenger transport, job pooling is often called shared rides or multi-load trips. Pooling can be prohibited on individual jobs when a customer requires a direct route. Learn more about job pooling.

Restrictions on which drivers can be assigned to a job. A required driver must complete the job, a preferred driver is assigned if possible, and a prohibited driver is never assigned. Learn more about job required drivers.

The amount of vehicle capacity currently in use at a given point in a route. The load fluctuates as stops are completed: pick-up stops increase it, delivery stops decrease it.

A mandatory job incurs a medium-level penalty when left unassigned, meaning it takes priority over all soft constraints. An optional job incurs a soft-level penalty when left unassigned and may be deferred when other soft constraints compete. Learn more about priority jobs and optional jobs.

The maximum amount of extra time a job’s stops can take compared to the direct travel time between them. Limits how long passengers or items spend in transit when jobs are pooled, ensuring they are not excessively delayed by other stops. In passenger transport, this limit is often called the maximum ride time. Learn more about maximum time burden.

A stop with a time window spanning multiple days that can be scheduled on any of those days. Contrast with non-movable stops, which have a single-day time window. Learn more about multi-day schedules and movable stops.

A pick-up where the client isn’t at the pick-up location when the driver arrives. Real-time planning replans the rest of the driver’s shift so the freed-up time goes to other jobs. Learn more about no-shows.

Transporting clients who don’t need emergency care to and from medical appointments, such as doctor visits, dialysis, or therapy. NEMT is a common use case for the Pick-up and Delivery Routing model. Learn more about use cases.

The percentage of stops completed within their time windows. In passenger transport, OTP is a key measure of service quality. The model plans every assigned stop within its time windows, so OTP is measured against actual arrival times and tracked in your own systems. The percentageStopsInSla and percentageJobsInSla KPIs measure stop SLAs, not time windows, so they aren’t the same as OTP. Learn more about time windows and opening hours.

Time a driver works beyond the preferred end of their shift. Overtime incurs a soft penalty, so Timefold assigns it only when it improves the overall plan, for example, when no other driver is available to complete a job. Learn more about overtime.

Locking a stop to its assigned driver and position in the route so it cannot be moved during re-planning. Useful when a driver is already en route to or working on a stop. Learn more about pinning stops.

The time period that will have jobs scheduled during a planning session. Learn more about planning windows.

A constraint that prevents two specific jobs from overlapping in the same driver’s route. Useful when passengers or items must not be transported together, for example when customers have requested not to travel with each other. Learn more about prohibited job combinations.

The ability to update a plan after it has been generated, to incorporate new jobs, cancellations, or driver availability changes while using the existing plan as a starting point. Learn more about real-time planning.

An API that quickly provides a range of time window options for a job when a customer is requesting service, without waiting for a full solve. Learn more about recommendations.

The process of minimizing the travel time or distance drivers spend driving between stops. By default, Timefold minimizes travel time. Learn more about route optimization.

A work period with start and end times when a driver is available for jobs. Learn more about shift hours and overtime.

Capabilities that drivers possess and jobs may require. Skills can have levels, allowing jobs to specify a minimum proficiency (for example, basic or advanced first-aid). Learn more about skills.

A commitment that a stop must be completed by a specific time. Stops that miss their SLA deadline incur a soft score penalty. Learn more about stop SLAs.

An urgent pick-up, typically of a medical specimen, that must be fitted around scheduled runs. STAT pick-ups are modeled as high-priority jobs with a stop SLA and added with real-time planning. Learn more about priority jobs.

A location within a job where the driver performs work, such as picking up or delivering an item. Stops must be completed in the order they are listed within a job. Learn more about time windows and opening hours.

The number of stops a driver completes per hour of shift time. Together with miles per stop and cost per stop, it measures route efficiency. Reducing travel, including deadhead miles, raises stops per hour and lowers cost per stop. The model doesn’t report these figures directly. You can calculate stops per hour from each shift’s itinerary, startTime, and endLocationArrivalTime in the output dataset, and distance per stop from the totalTravelDistanceMeters and totalAssignedStops KPIs. Learn more about model output and KPIs.

Labels applied to driver shifts to model binary requirements such as certifications or vehicle attributes. In passenger transport, tags commonly represent the vehicle type a client needs, for example, a wheelchair-accessible van or a stretcher vehicle. Jobs can require specific tags, restricting which drivers can be assigned. Learn more about tags.

The period during which a stop can be serviced. If the driver arrives before the time window opens, they must wait. If they cannot arrive before the time window closes, the stop (and potentially the whole job) may be left unassigned. Learn more about time windows and opening hours.

A return trip with no set pick-up time, where the client calls when they’re ready to be picked up, for example, after a medical appointment. Will-call trips are added to the plan with real-time planning. Learn more about real-time planning.