From 2015, Crossrail Ltd managed the railway programme against an aspirational plan designed to push suppliers along. It presented that plan as the critical path to completion. The National Audit Office’s 2019 report found that a detailed, realistic plan built from the work itself did not begin until late 2018.
The central London section had been due to open in December 2018. Crossrail Ltd said that August it could not meet the date. The plan had not adequately accounted for dependencies across the programme, so its apparent route to opening gave a poor test of the work and risk still ahead, according to the NAO’s account.
A date can be a target that presses work forward. It also has to survive contact with what the team has finished. Crossrail shows what happens when the first job of a plan is allowed to stand in for the second.
The reset repeated the assumption
The NAO’s 2021 update found that the revised schedule and budget agreed in April 2019 were unachievable. Crossrail Ltd and its sponsors had understood the programme to be nearer completion than it was. The new management team had to start largely from scratch. Contractors met around 30% of their milestones on average through 2019 and early 2020, while the plan continued to assume a higher rate.
The COVID-19 pandemic then added cost and delay. Crossrail Ltd used the construction pause to improve its planning with contractors. Around 90% of milestones were met on average between September 2020 and April 2021, the NAO reported. Those periods involved different conditions, and the figures do not prove that a better plan alone caused the change. They do show why a new date needs to be checked against work completed under the conditions in which the team is working.
For a smaller team, compare the milestones due in a recent, relevant period with those accepted on time. Look at the missed ones and the new work discovered. Then ask whether the proposed finish date assumes a pace the team has yet to achieve. A percentage from last month cannot simply be multiplied across a different mix of tasks, but it can expose an estimate that depends on an unexplained jump in delivery.
Completion includes the joins
Crossrail Ltd and its contractors also underestimated the work of bringing separate systems together and proving that the railway was safe and operational. The 2021 report described around 500,000 physical and digital assets that had to be assured individually and in combination, with operating manuals and processes for the railway as a whole.
This is the part of an estimate that a list of finished components can miss. A team may have produced each item promised while the person taking over still needs access, tests, instructions and time to inspect the combined result. Before accepting a finish date, make that work visible, including time to correct what testing finds.
The person who can change the promise
The Department for Transport and Transport for London set Crossrail’s scope, budget and timetable as sponsors. Crossrail Ltd delivered the programme, but the 2019 report found that the sponsors’ high degree of delegated autonomy left them with few effective contractual levers late in delivery. In October 2020 the governance was changed, giving TfL’s Transport Commissioner ultimate accountability for completing it, according to the 2021 update.
On a smaller project, the person managing delivery can show why the date is no longer credible and recommend a course. The people with authority over the promise must decide on a changed date, scope, funding or acceptance condition. If that route is unclear, identifying who can make the decision is itself part of the remaining work. A delivery plan cannot substitute for their agreement.
Crossrail’s first passenger service did not mean every task was complete. In 2021 the NAO reported that some residual work had been agreed for after opening, while its eventual cost and even its necessity were uncertain. It recommended a handover plan specifying what was deferred, who was responsible and how progress would be monitored. TfL records that the central section opened on 24 May 2022.
At the next progress review, put the promised date beside the work actually accepted in a comparable recent period and the tests still to be run. If the three do not fit, take the discrepancy to the person who can change the promise, with the work and decision still outstanding in view.