From Black Box to Spreadsheet: Engineering Certainty into Construction Project Control
Project management theory has advanced tremendously over the last generation. In practice, however, it has often fostered overconfidence and contributed to failure. Construction contracts frequently require submission of a project schedule in MS Project or Primavera, yet these documents often become little more than a paper formality with limited ongoing use.
Both tools are powerful and general-purpose, offering far more robust scheduling capabilities (critical-path analysis, resource levelling, baselines, etc.) than the ubiquitous Excel spreadsheet. With the arrival of Copilot in Excel—already present on the desks of most field engineers—the practical advantage has begun to shift.
Smart construction managers have long used the Last Planner System (scheduling with reliable, bottom-up commitments from the people closest to the work) and its associated lookahead plans — collaborative practices formalised in the 1990s that have since become central to lean construction. The main practical difficulty was keeping pace with resource-requirement calculations while managing a schedule with complex dependencies and constraints. Excel with Copilot now resolves this challenge in a way that aligns with the last planner’s ever-changing commitments.
While it is not entirely accurate to describe MSP and Primavera as black boxes—given their capacity to combine and process multiple data sources into simplified visual representations—their high cost and steep learning curve firmly establish them as top-down tools. They excel in baselining and business modelling, but prove less effective for detailed look-ahead resource calculations and for integrating the complexities of the Last Planner System.
Excel is already a fixture on most construction site managers' desks and can be readily scaled to support the Last Planner System. When paired with Copilot, it changes the game for construction project control.
Resource Requirement Calculations
Based on DAR
https://cpwd.gov.in/WriteReadData/Publication/CPWD_DAR_Vol_I_14092023-Civil.pdf,
calculate the LABOUR resources required for 100 Cum of item no 6.1.1 on Page
307
Scheduling
Prepare a daily
requirement for each labour resource and list them day-wise in the Resource
sheet. Use days in the rows (Day-1, Day-2, etc.) and a column for each
resource. Complete the entire work within 28 days. Item6.1.1 shall be completed
by Day 15, Item13.3.2 would be taken up from Day 7 to day 22. Item13.47.1 would
be taken up from day 15 to day 28. List in such a way so that when I change the
project or item duration, the day wise resource requirement is recalculated.
Assume linear distribution i.e. equal work is done each day within the
schedule. Lock all calculated and constant cells of sheets, leaving only
variables i.e. quantity value and schedule variables open. Highlight such
variable cells. Protect sheets without passwords.
You can download the workbook here and use it to learn and test yourself by creating a detailed practical solution for your next look-ahead schedule from the Last Planner inputs. Prepare a daily
requirement for each labour resource and list them day-wise in the Resource
sheet. Use days in the rows (Day-1, Day-2, etc.) and a column for each
resource. Complete the entire work within 28 days. Item6.1.1 shall be completed
by Day 15, Item13.3.2 would be taken up from Day 7 to day 22. Item13.47.1 would
be taken up from day 15 to day 28. List in such a way so that when I change the
project or item duration, the day wise resource requirement is recalculated.
Assume linear distribution i.e. equal work is done each day within the
schedule. Lock all calculated and constant cells of sheets, leaving only
variables i.e. quantity value and schedule variables open. Highlight such
variable cells. Protect sheets without passwords.

Treating AI as some form of voodoo will get us nowhere. Yet this attitude persists across sectors, creating widespread confusion. I see Ashutosh as a pioneer—an enlightened engineer, as I have always known him since our days together at the College of Technology, G.B. Pant University, in the 1970s. He approaches AI as an enabler, with clarity about the task it must serve. His effective use of Copilot to turn Excel into a highly adaptable scheduling tool is a brilliant example. It demonstrates how specific, precisely tailored prompts, guided by sound engineering judgement, can bring substantial gains in efficiency. May the tribe of Ashutosh increase.
ReplyDelete