Gantt Chart

The foundational axes of the chart. The vertical axis (Y-axis) lists all project tasks or work packages. The horizontal axis (X-axis) represents the project timeline, scaled in appropriate units (days, weeks, months). This creates a matrix where each task's duration is plotted as a bar along the timeline, providing an immediate visual overview of the schedule's scope and timeframe.

The primary visual element. Each task is represented by a horizontal bar. The bar's left edge indicates the planned start date, and its right edge indicates the planned finish date. The bar's length is directly proportional to the task's duration. In advanced scheduling for heterogeneous fleets, bars can be color-coded or patterned to represent:

• Different resource types (e.g., autonomous mobile robot vs. manual forklift).
• Task priority levels.
• Current status (e.g., on-track, delayed, completed).

Lines or arrows connecting task bars that define precedence constraints. These are critical for modeling workflows where one task cannot start until another finishes (Finish-to-Start). Common dependency types include:

• Finish-to-Start (FS): The most common; successor starts after predecessor finishes.
• Start-to-Start (SS): Successor can start once predecessor starts.
• Finish-to-Finish (FF): Successor finishes when predecessor finishes.
• Start-to-Finish (SF): Rare; successor finishes when predecessor starts. In fleet orchestration, these model the sequence for multi-agent tasks, like an AMR unloading before a manual vehicle can transport.

Significant, zero-duration events or checkpoints in the project lifecycle, represented by diamonds or triangles on the timeline. They mark critical achievements, such as:

• Project phase completions.
• Key deliverable approvals.
• External dependency handoffs.
• Major system go-live events. Milestones are used to track overall progress and synchronize the efforts of different teams or agent groups within a coordinated schedule.

The mapping of tasks to specific resources (agents, machines, personnel). In a basic Gantt chart, this may be noted in the task list. In sophisticated spatial-temporal scheduling systems, this is integrated to show:

• Which specific autonomous mobile robot or vehicle is assigned to a task bar.
• Resource overallocation (conflicts where one agent is scheduled for two simultaneous tasks).
• Utilization rates across the heterogeneous fleet. This component is essential for ensuring the schedule is not just temporally feasible but also resource-feasible.

Mechanisms for monitoring execution against the plan. This often involves:

• A baseline schedule (the original plan), often shown as a faint outline or a different-colored bar.
• Percent-complete shading within task bars, visually indicating progress (e.g., 60% of the bar is filled).
• A vertical 'Today' line that moves with the current date, providing an instant visual comparison between planned work and actual progress. This allows for real-time schedule robustness assessment and triggers online scheduling or real-time replanning when deviations occur.

A live Gantt Chart serves as the primary operational dashboard, visualizing the current state and planned schedule of every agent in the fleet. Each horizontal bar represents an agent (AMR, AGV, or manual vehicle), with color-coded segments showing:

• Active Task Execution (e.g., picking, transporting)
• Travel Time between locations
• Idle/Wait Time (often indicating a bottleneck)
• Scheduled Maintenance or Charging This provides at-a-glance situational awareness for human supervisors, showing which agents are on schedule and which are delayed.

Gantt Charts excel at revealing precedence constraints and resource conflicts that are critical in warehouse workflows. For example:

• A packing station cannot start until a fetching robot delivers the required items (precedence).
• Two autonomous mobile robots cannot occupy the same narrow aisle simultaneously (spatial-temporal conflict). Bars are linked with lines to show dependencies, and overlapping bars on shared resources (like a single unload dock) are instantly visible, allowing planners to manually or automatically adjust schedules to resolve deadlocks.

For electric fleets, Gantt Charts integrate energy constraints directly into the schedule. Bars include distinct segments for:

• Operational runtime at current battery level
• Travel to charging station
• Charging duration (which may vary based on depth of discharge)
• Return to active duty This allows schedulers to proactively plan charging cycles, ensuring high-priority tasks are not assigned to agents nearing low battery, thereby maintaining continuous fleet uptime and adhering to capacity constraints of the electrical infrastructure.

When disruptions occur—like an agent failure or a high-priority rush order—the Gantt Chart becomes an interactive tool for online scheduling. Planners can:

• Drag-and-drop task bars to reassign work.
• Visualize the makespan impact of inserting a new job.
• Simulate different priority-based routing rules. Advanced orchestration platforms use this visual model as the front-end for Model Predictive Control (MPC) systems, where the underlying solver continuously re-optimizes the schedule over a receding horizon, with changes reflected in real-time on the chart.

The Gantt Chart is the human-readable output of complex backend optimization algorithms. It represents a feasible solution to a Mixed-Integer Programming (MIP) or Constraint Programming (CP) model that considers:

• Vehicle Routing Problem (VRP) with time windows
• Job Shop Scheduling for stations
• Load balancing across the fleet Solvers like CP-SAT or Gurobi generate the optimal task sequence and timing, which is then rendered as the Gantt visualization. The chart can also display the optimality gap during solver execution.

Historical Gantt Charts are aggregated for post-shift analysis, transforming schedule data into performance insights. By analyzing patterns in bar lengths and gaps, operations researchers can identify:

• Chronic bottleneck resources (stations with consistent waiting queues).
• Inefficient agent utilization (excessive idle time).
• Schedule robustness to common delays.
• Average task duration variance versus plan. This data feeds back into the Digital Twin for discrete-event simulation, used to test new scheduling policies and improve schedule robustness before live deployment.