Critical Path Method (CPM) & Network Logic

Critical Path Method (CPM) & Network Logic

Quick Answer: The Critical Path Method (CPM) is a mathematical scheduling technique that models project activities as a deterministic network of dependent nodes. By executing a forward pass, the scheduler identifies the earliest possible start and finish dates for each task; a backward pass reveals the latest dates an activity can occur without extending project completion. The Critical Path is the longest continuous sequence of activities through the network and possesses zero total float. Any delay to an activity on the critical path directly delays the overall project completion date.

---

CPM Fundamentals & Precedence Network Logic

Modern AEC scheduling primarily utilizes the Precedence Diagramming Method (PDM), also commonly termed Activity-on-Node (AON). In an AON network, discrete project activities are represented by rectangular boxes (nodes), and logical dependencies are depicted by connecting arrows.

The Standard AON Node Layout

Each node in an AON network diagram contains six essential pieces of schedule data:

+-----------------------+----------+-----------------------+
| Early Start (ES)      | Duration | Early Finish (EF)     |
+-----------------------+----------+-----------------------+
|             Activity Name & Description                  |
+-----------------------+----------+-----------------------+
| Late Start (LS)       | Float    | Late Finish (LF)      |
+-----------------------+----------+-----------------------+

Logical Dependency Relationships

Dependencies define the sequential logic between predecessor and successor activities. In architectural production and construction scheduling, four precedence relationships exist:

1. Finish-to-Start (FS): The successor activity cannot start until the predecessor activity finishes. This is by far the most common logic tie in architectural practice (e.g., the structural framing plans cannot start until the architectural column grid is finalized).
2. Start-to-Start (SS): The successor activity cannot start until the predecessor activity has started. This allows overlapping workflows (e.g., mechanical duct modeling can start once architectural core modeling starts).
3. Finish-to-Finish (FF): The successor activity cannot finish until the predecessor activity finishes (e.g., specification writing cannot finish until construction drawings finish).
4. Start-to-Finish (SF): The successor activity cannot finish until the predecessor activity starts. This relationship is exceptionally rare in architectural practice and is generally avoided.

Dependencies may also incorporate Lead (negative lag that accelerates successor work by starting it prior to predecessor completion) or Lag (mandatory delay between activities, such as requiring a 7-day curing lag after pouring concrete before structural steel erection commences).

---

Mathematical Mechanics: Forward and Backward Passes

Determining project duration and critical path requires executing two sequential mathematical sweeps through the network: the forward pass and the backward pass.

1. The Forward Pass (Early Dates)

The forward pass moves chronologically from the project start node to the final completion node. It establishes the Early Start (ES) and Early Finish (EF) for every activity.

• The initial activity starts at time zero: $ES_1 = 0$.
• The early finish is calculated by adding activity duration:

$$EF = ES + \text{Duration}$$

• Merge Node Rule: When multiple predecessor activities converge upon a single successor activity, the successor's Early Start is governed by the maximum Early Finish among all immediate predecessors:

$$ES_j = \max(EF_i) \quad \text{for all immediate predecessors } i$$

2. The Backward Pass (Late Dates)

The backward pass moves in reverse from the final project completion date back to the start node. It establishes the Late Finish (LF) and Late Start (LS)—the latest possible dates an activity can be completed or initiated without delaying the overall project completion date.

• For the final project activity, Late Finish equals its Early Finish: $LF_{\text{final}} = EF_{\text{final}}$.
• The late start is calculated by subtracting activity duration:

$$LS = LF - \text{Duration}$$

• Burst Node Rule: When moving backward, if an activity is followed by multiple successors, its Late Finish is governed by the minimum Late Start among all immediate successors:

$$LF_i = \min(LS_j) \quad \text{for all immediate successors } j$$

---

Float Mechanics: Total Float vs. Free Float

Float (or slack) represents the scheduling leeway available to an activity. On the ARE 5.0 Project Management exam, distinguishing between Total Float and Free Float is critical:

Total Float (TF)

Total Float is the total amount of time that an activity can be delayed from its Early Start date without delaying the project's overall contractual completion date. It reflects the flexibility of the activity relative to the entire project schedule:

$$\text{Total Float} = LS - ES = LF - EF$$

• If $\text{Total Float} = 0$, the activity is Critical. Any delay to this activity will cause an identical day-for-day delay to the project completion date.
• If $\text{Total Float} > 0$, the activity is Non-Critical and possesses schedule leeway.
• If $\text{Total Float} < 0$, the project is facing Negative Float, meaning the current schedule sequence cannot meet the mandated contract deadline unless activities are accelerated or compressed.

Free Float (FF)

Free Float is the amount of time an activity can be delayed without delaying the Early Start of any immediate successor activity. Free float represents localized slack that does not affect downstream activities:

$$\text{Free Float} = \min(ES_{\text{successors}}) - EF$$

Notice that in the unconstrained example, Free Float does not exceed Total Float ($FF \le TF$). While an activity may have several days of Total Float, its Free Float will be zero if any of its successors must begin immediately upon its early completion.

---

Step-by-Step Worked Example: 8-Activity Architectural Network

Consider an architectural team executing the Schematic Design phase for a specialized municipal facility. The project consists of eight discrete activities (A through H) with established dependencies and durations in calendar days:

Step-by-Step Mathematical Walkthrough

1. Forward Pass Calculation:

   • Activity A: Starts at day 0. $EF = 0 + 4 = 4$.
   • Activity B: Predecessor is A ($EF = 4$). $ES = 4$; $EF = 4 + 6 = 10$.
   • Activity C: Predecessor is A ($EF = 4$). $ES = 4$; $EF = 4 + 10 = 14$.
   • Activity D: Predecessor is C ($EF = 14$). $ES = 14$; $EF = 14 + 8 = 22$.
   • Activity E: Predecessors are B ($EF = 10$) and C ($EF = 14$). Applying the merge rule: $ES = \max(10, 14) = 14$. $EF = 14 + 5 = 19$.
   • Activity F: Predecessor is E ($EF = 19$). $ES = 19$; $EF = 19 + 4 = 23$.
   • Activity G: Predecessors are D ($EF = 22$) and F ($EF = 23$). Applying the merge rule: $ES = \max(22, 23) = 23$. $EF = 23 + 7 = 30$.
   • Activity H: Predecessor is G ($EF = 30$). $ES = 30$; $EF = 30 + 3 = 33$.
   • Overall Project Duration: 33 Days.

2. Backward Pass Calculation:

   • Activity H: $LF = 33$; $LS = 33 - 3 = 30$.
   • Activity G: Successor is H ($LS = 30$). $LF = 30$; $LS = 30 - 7 = 23$.
   • Activity D: Successor is G ($LS = 23$). $LF = 23$; $LS = 23 - 8 = 15$.
   • Activity F: Successor is G ($LS = 23$). $LF = 23$; $LS = 23 - 4 = 19$.
   • Activity E: Successor is F ($LS = 19$). $LF = 19$; $LS = 19 - 5 = 14$.
   • Activity B: Successor is E ($LS = 14$). $LF = 14$; $LS = 14 - 6 = 8$.
   • Activity C: Followed by D ($LS = 15$) and E ($LS = 14$). Applying the burst rule: $LF = \min(15, 14) = 14$. $LS = 14 - 10 = 4$.
   • Activity A: Followed by B ($LS = 8$) and C ($LS = 4$). Applying the burst rule: $LF = \min(8, 4) = 4$. $LS = 4 - 4 = 0$.

3. Network Path Analysis & Float Verification:

   • Path 1: A $\rightarrow$ B $\rightarrow$ E $\rightarrow$ F $\rightarrow$ G $\rightarrow$ H = $4 + 6 + 5 + 4 + 7 + 3 = 29\text{ days}$. ($TF = 33 - 29 = 4\text{ days}$ on Activity B).
   • Path 2: A $\rightarrow$ C $\rightarrow$ D $\rightarrow$ G $\rightarrow$ H = $4 + 10 + 8 + 7 + 3 = 32\text{ days}$. ($TF = 33 - 32 = 1\text{ day}$ on Activity D).
   • Path 3: A $\rightarrow$ C $\rightarrow$ E $\rightarrow$ F $\rightarrow$ G $\rightarrow$ H = $4 + 10 + 5 + 4 + 7 + 3 = 33\text{ days}$. This is the longest path through the network.
   • The Critical Path is A $\rightarrow$ C $\rightarrow$ E $\rightarrow$ F $\rightarrow$ G $\rightarrow$ H with total float of 0 days.

---

Schedule Compression Techniques: Fast-Tracking vs. Crashing

When a project falls behind schedule or an owner demands an earlier completion date, the project manager must compress the project schedule. Two primary compression techniques exist, each carrying fundamentally different financial and risk profiles:

Fast-Tracking (Parallel Processing)

Fast-tracking involves taking activities that were originally planned to occur in sequence and reconfiguring them to execute concurrently in parallel.

• Example in Architectural Practice: Beginning foundation and structural concrete construction documents while architectural schematic layouts are still being finalized with the client.
• Cost Impact: Fast-tracking does not directly increase labor costs or require overtime fees initially.
• Risk Impact: Extremely High Risk of Rework. If schematic layouts change after structural foundation engineering has begun, completed drawings must be scrapped and redrawn. Fast-tracking frequently causes coordination errors, RFIs, and costly change orders during construction.

Crashing (Resource Injection)

Crashing compresses the schedule by injecting additional resources onto critical path activities to shorten their duration.

• Example in Architectural Practice: Assigning two additional senior drafters to work mandatory overtime on Activity G (Integrated BIM Coordination) or hiring an outside drafting service to expedite drawing production.
• Cost Impact: Substantial Direct Cost Increase. Crashing requires paying premium overtime rates, expediting fees, or additional staff wages. Furthermore, crashing is subject to the economic law of diminishing marginal returns—overcrowding a task can lead to communication inefficiencies and reduced individual productivity.
• Risk Impact: Lower technical risk than fast-tracking because activity sequence and dependencies are preserved, but high risk of budget overruns and staff burnout.
• Critical Rule: Crashing must only be applied to activities on the Critical Path. Adding resources to an activity with positive float (such as Activity B or Activity D in our example) increases project costs without accelerating the overall project completion date by even one day.

The forward/backward-pass examples assume a day-zero origin, finish-to-start dependencies, no lags, and no imposed finish constraint. The usual nonnegative float relationships apply to this network; imposed dates can produce negative total float. Fast-tracking can increase rework and cost, even though it changes overlap rather than simply adding resources. Recalculate the network after an approved change.