The Virtual Building Group has been developing a Scheduling Dashboard to help us achieve operational excellence in our construction scheduling. With the Dashboard we are able to monitor and track the health of every project’s schedule, ensure that our people know and use proper scheduling techniques, and hold people accountable to company standards.
The Dashboard is a web application where schedules created in P6, Synchro, or even Microsoft Project can be uploaded for project managers to self-audit their schedules and for VBG schedulers and operational leadership to review project status. The application shows the schedule health of all Beck projects from a very high level all the way down to individual tasks. The scheduling data is tracked over time, so we can: view changes in contract completion dates, days gained/lost, number of scheduled activities, and maintain baselines, duration changes and proper logic. This helps us be proactive in keeping any individual job on schedule and helps us manage our projects across the company.
The main Dashboard page (above) displays a list of projects with each project’s most recent schedule health score and a timeline of calendar days elapsed and remaining in the contract period. Clicking the score’s details link takes you to a table of schedules with basic date/milestone/task information and the final score for each schedule uploaded (below). Some of the cells are pre-requisites for achieving a good final score. An example of a pre-requisite is a maximum of 31 days between schedule updates. Contractually we are supposed to update the schedule if we ever are in need to request additional General Conditions due to Owner changes or inclement weather that extends the duration of the project completion date. If we are not updating the schedule, we will most likely not receive money that is owed to us.
For historical reasons, scores range from 0 to 100, with 0 meaning no issues and 100 indicating significant problems. Users can dive down to a page containing further details to look at specific audit areas where their schedule might not adhere to company policy. The Dashboard checks things such as
- Checking of task-to-task sequential logic.
- Inclusion/exclusion of holidays
- Task duration
- Lag between tasks
- Task constraints
One user requested the ability to compare a revised schedule to a baseline schedule, which we now do in table form (below). Rows and cells of the table are color-coded to indicate tasks which differ by appearing only in one schedule and not the other, or by appearing in both and having date or percentage completion differences.
Another user request was for a way to link back to the original scheduling software (e.g. Synchro). Although we couldn’t do that, we could recreate a Gantt chart on which we visually highlight schedule problems (below).
Currently, all projects upload their schedules to the Leader’s Top Ten database at the end of each month. Unfortunately, when these schedules are in PDF format there is no quick way of determining if tasks are correctly linked, holidays properly accounted for, or even whether or not the project has been updated. This leaves Beck exposed to risk, and if problems are occurring with the progress of a job it is almost always too late to correct issues in time to meet the contractual project completion date without exhausting our resources. To address these risks, project schedules will be uploaded to the Dashboard instead of Leader’s Top Ten starting in January 2015. A link will be provided at that time.
Other folks are taking notice of our efforts. When we mention the Scheduling Dashboard at conferences, people say they wish they had this capability. We recently won the AGC (Washington DC chapter) Innovation in Technology Award sponsored by the Washington Business Journal, given for the “best use of technology as a solution to an AEC industry issue or problem”.
We look forward to expanded use and improvements to the tool in 2015. Here’s a list of enhancements already in the pipeline:
- Auto-generating schedule templates based on project type, expected project phases, etc. selected by the user.
- Cross-referencing job cost information
- Weather delay analysis
- Machine learning/statistical analysis of what constitutes a “good” schedule