Lean Construction Ireland Annual Book of Cases 2021 43 If the recorded actual outputs varied from the planned outputs, we were able to react tomake the necessary adjustments by, for example, increasing resources, resequencing, or removing blockers, and then updating the Pull Plan to achieve the target date. A prime example of this was evident during a Kaizen event that occurred during the steel frame installation. It resulted in the final torqueing of the bolts being completed on windy days when the crane was not operational so that time wasn’t lost as a result of the crane being winded off. This allowed the crews to be more productive when the crane wasn’t operational. While many benefits were realised from this data collection approach, as the project progressed and more trades came on board, a reduction in the level of detail, quality, and consistency in the data collection was noticed. Figure 2. Six-Week Pull Plan New Lean Initiative Define: During this stage of this Lean continuous improvement project, we discussed the current data collection process and why there were varying levels of detail and consistency. It was accepted that the then data collection strategy focused on lengthy non-valueadding data inputting with no owner assigned to the information flow. For example, the data was input both on the whiteboard (for communication at the weekly Pull Plan session) and the Master Excel Tracker (which we used for overall activity analysis) before being fed back into the Master Programme at the end of each week. There was an underestimation of the time required to collect and input the data into the various trackers before filing the information away.There was also the added risk of erroneously wiping the board and losing the data before it was captured on the Master ExcelTracker. Using Lean tools such as drawing theAs-Is Process Map and SIPOC, we determined areas for improvement and settled on our goal statement: Improving the planning and tracking of progress by streamlining the data collection process and whilst also reducing non-value-adding work. Measure: When reviewing the collected data, we noted several instances of incomplete information.With the complete data, we were able to compile a Pareto analysis of the reasons for delays encountered during the installation period of the key trades.This graphical analysis helped create a picture of what ‘pain’ we were suffering in the process.While this information is good to capture, it focuses more on a reactive approach than being proactive. Figure 4. Pareto Chart Analyse: Here we listed the probable reasons for incomplete data collection and used the Cause & Effect Matrix and FiveWhys to get to the root cause. Primarily, the reasons were: • Not understanding the benefits of data collection; and • Not being prompted to input complete information. Improve: As a group, we brainstormed possible solutions and depicted the improvement ideas on a PICK chart to gauge the potential payoff and level of difficulty of implementation.A quick and easy solution that was deemed to have a high level of pay-off was an online form that would be filled out at the end of each day by the Package Manager on-site on their mobile device. It would quantify the actual work completed that day, plus any issues (‘pain’) encountered, as well as planned work for the following day.This form could be accessed easily via a QR code positioned at the site exit. We also applied Poka-Yoke for error proofing by using compulsory fields to prevent non-responses and incomplete data collection.Once submitted, the data would be instantly available for several parties to view, improving levels of communication since several members of Figure 3. Daily Productivity Board Case 11