How to Build a Weekly Construction Crew Schedule with AI Foreman

Manual weekly scheduling for construction and field-service crews is a complex puzzle. Balancing skills, travel, and customer communication often leads to inefficiencies and missed opportunities. You need your best people on work orders without wasting time or leaving customers in the dark. WorkSkedge's Advanced Dispatch and Advanced Scheduling tools, combined with the AI Foreman, tackle these challenges.

These tools provide a clear view of your next 7 to 14 days, help draft optimal crew assignments, manage complex work order requirements like travel, and automate customer communications. This guide explains how to use the AI Foreman to build, refine, and confirm your weekly schedule, ensuring work orders run smoothly and resources are used effectively.

Leveraging AI Foreman for Weekly Crew Scheduling

Effective weekly scheduling starts with understanding upcoming work orders and crew availability. WorkSkedge's Advanced Dispatch must be enabled for your company to use its tools, including the AI Foreman. You can check its status in your company settings.

Once enabled, the Advanced Scheduling workbench displays a crew-by-day grid for up to 14 days. This grid shows current assignments, available time (slack), and unscheduled work orders. The AI Foreman, referred to as Eric in internal documentation, reads this schedule view, along with your list of unscheduled work orders and all your company's scheduling rules.

The AI Foreman drafts schedule plans. It does not apply them directly. These drafts require human review and commitment, ensuring your control over the final schedule. This process allows the AI to handle drafting while your expertise and final approval remain central.

For customer communication, the AI Foreman can propose contacts. These proposals require human approval if your communication settings require it. This ensures you control outgoing messages. The system also helps determine the best contact method, whether automatic text or call only.

To get customer confirmations, the AI Foreman can offer plan dates for entire plans or for individual days. Customer responses are recorded. If a live work order needs to move, a similar process initiates, creating a draft for the change.

AI-Powered Draft Scheduling for Optimal Crew Deployment

Starting your weekly schedule with the AI Foreman provides a smart draft that considers all your rules and crew capabilities. The AI Foreman analyzes your unscheduled work orders to identify those needing a date or a crew. It then proposes a comprehensive draft plan for the coming 7 or 14 days.

When drafting, the AI Foreman considers your company's scheduling rules, including crew pairing rules and crew composition rules. Crew pairing rules dictate who must work together and who must never be paired. Crew composition rules define the ideal crew for a specific work order type, including who can lead and the required mix of roles or groups. These composition rules are advisory, producing warnings rather than hard blocks, but they influence the AI's suggestions.

The AI Foreman also considers crew performance, factoring in on-time starts, report submission rates, and overall scores. To fill an empty seat, it ranks potential crew members based on availability, required skills, and continuity. It can also evaluate a proposed crew against your rules.

After considering these factors, the AI Foreman builds a draft schedule. This draft is a proposed set of bookings, not a live schedule change. It is designed for your review and refinement, ensuring the AI's suggestions align with your operational needs before anything goes live.

Managing Complex Scheduling Requirements: Travel and Overnight Shifts

Construction and field service work often involves work orders requiring travel or overnight shifts. These add complexity that the AI Foreman handles, provided the necessary information is explicitly set.

For work orders requiring travel, the system relies on a flag. A work order is considered away work only when explicitly marked as "yes" by the scheduler. The system never infers travel from distance or work order length. For older work orders, if marked "auto," they count as away work only if they involve overnight stays.

When the AI Foreman reviews unscheduled work orders, it identifies those with travel required, a travel label, nights away, and travel hours. Your company's scheduling rules also provide company-wide settings like default travel hours (typically 4 hours round-trip) and maximum consecutive work days (default 6 days). While travel hours is a reference for expected round-trip time and does not alter scheduled start/end times, the maximum consecutive work days is enforced. The system will block anyone whose consecutive work days would exceed this limit.

Travel is booked as its own shift. A day with both driving and work will have two distinct shifts: a travel shift and a work shift. Travel shifts are marked as "travel" and can be "outbound" or "return." This distinction is crucial for timesheets and customer reminders, as customers are only informed about on-site work.

Overnight shifts, which cross midnight, also require explicit handling. When proposing such a shift, you must set the "overnight" flag on the plan item whenever the end time is at or before the start time (e.g., 22:00 to 06:00). Omitting this flag for a cross-midnight shift, or setting it for a non-cross-midnight shift, results in a hard rejection. On commit, an overnight flagged day becomes two linked bookings, sharing a shift ID and marked as overnight, with a shift segment as evening or morning. Work-hours checks are skipped for overnight items. When reading existing overnight shifts, always group by shift ID to understand them as a single continuous shift.

Reviewing and Refining AI-Generated Draft Schedules

Once the AI Foreman generates draft plans, your role is to review and refine them. These drafts are scenarios, not live schedule changes, offering flexibility for adjustments before finalization. You can view all draft schedule plans, which shows their status: draft, awaiting customer, approved, committed, rejected, or cancelled.

To examine a specific draft, you can view every detail: work order, day, times, assigned employee, site lead flag, and the AI's rationale. This allows you to scrutinize suggestions and understand the reasoning.

Before seeking customer approval, the system identifies any clashes in uncommitted draft items against the live schedule. This includes existing assignments, time off, or days without a crew. This helps catch potential problems early.

Drafts are dynamic. If a customer requests a change, you can update an existing draft's dates. The customer receives one communication for the entire draft, so their reply applies to all work orders within it. The system automatically recalculates the draft's window and derived name to reflect these changes.

The AI Foreman is constantly aware of changes to its drafts. Any edit to a draft's dates or crew, rejection, deletion, or commitment triggers an event. This event includes flags indicating if the draft is still active, what kind of change occurred (window, crew, or items), and if any AI runs were cancelled. This ensures the AI never works from outdated information.

When reviewing, consider company working days and holidays. Your scheduling rules provide company holidays, and the schedule view flags each day as a working day, a holiday, and the holiday name. The system hard-blocks scheduling on non-working days and holidays, so the AI Foreman's drafts respect these rules.

If manual adjustments are needed, you can add work orders to an existing draft from the grid, the unscheduled queue, or the available-crew panel. The draft's name is derived from the customer and its window, but you can override it. The panel summarizes the draft, showing customer, window, work order/day/crew counts, total hours, and who built it. Any day without a crew will clearly state "No crew assigned yet."

Automating Customer Communication and Confirmations

Once you have a solid draft schedule, the next step is customer buy-in. WorkSkedge automates much of this communication, with your oversight. The AI Foreman can propose customer contacts, but these are staged for your approval, especially if your communication settings require it. No message goes out without your explicit approval.

To initiate the booking loop, you can offer plan dates for an entire draft plan or for specific days. This parks the plan in "awaiting customer" status. The system determines the best contact method, which returns whether to use automatic text or call only, and the best textable phone number. If a customer is call-only or texting is disabled, the system opens a call request instead of sending a text.

WorkSkedge's customer texting settings, found under Customer texting in the gear menu on Advanced Scheduling, govern how these messages are sent. These settings include quiet hours (no texts delivered during this window), opt-outs (customers replying STOP are never texted again), and maximum rounds of text offers (default 2 rounds). If a customer uses up their offer rounds or replies "CALL," the day becomes a call request and is parked as "call requested," escalating to a human dispatcher.

Customers can reply to offered slots by text, or you can manually record their answer for a whole plan or for a single day. An "accept" outcome confirms the dates, moving the plan day to "confirmed" status. If a customer picks a slot that has since been taken, the system replies that the time is unavailable and offers fresh options.

For existing bookings that need to be rescheduled, moving the live assignment into the same booking loop creates a one-day draft plan that replaces the original, and by default, texts the customer the proposed new dates. The original booking remains on the schedule until the replacement draft is committed.

All customer messages, including replies and delivery statuses, are visible on the work order's Scheduling tab under "Customer messages." This tab also shows the work order's activity timeline and any reported disruptions, providing a complete communication and scheduling history.

Finalizing Schedules and Handling Adjustments

Once your draft plan is reviewed, refined, and customer-confirmed, the final step is to commit it to the live schedule. This is a human action; the AI Foreman only drafts, it never applies. Committing a draft plan marks it as "approved" server-side, writes every item to the real schedule, moves associated work orders to "assigned" status, and notifies the crew.

Before committing, the system performs a final conflict check on every item, as the live schedule may have changed since the draft was created. Any items that now conflict are reported back and left uncommitted, allowing you to address them. Committed items are marked as written and will not be re-applied if you commit the plan again.

After a plan is committed, any subsequent changes to the schedule, whether manual or AI-assisted, follow a similar process. If a live work order needs to be moved, a new draft plan is created for the change, which then goes through the customer confirmation loop, and the customer's answer is recorded.

Disruptions are common in construction. While the AI Foreman helps plan, handling disruptions often requires direct action. If a crew member is marked sick, their work orders are flagged for recovery, but nothing moves automatically. You choose the fix. All disruptions, customer messages, and scheduling activity for a specific work order are logged on its Work order Scheduling tab, providing a comprehensive history for review and auditing.

WorkSkedge ensures that even after a schedule is live, you have the tools to manage changes, communicate effectively with customers, and keep operations running smoothly. The AI Foreman is a powerful assistant, but final control and decision-making always remain with you, the operator.