The Potential for Artificial Intelligence to Revolutionise Project Management

by Daniel Birch, senior manager and Neal Murray, manager, PwC programme and project management / 2/8/2018 10:04:59 AM

Artificial Intelligence (AI) has the potential to revolutionise so many aspects of how we live and work, including many opportunities in project management.  AI tools can find patterns in large current and historical datasets of project information that could help project management professionals shape plans, identify and resolve risks and issues, and change how they do status reporting. 

With news reports and the Standish Group’s annual CHAOS study continuing to highlight the significant rate of failure within large projects like IT implementations and infrastructure development, now is the time for the project management profession to start using big data and AI to improve outcomes and increase the likelihood that projects will be successful.​

Information overload in project management

According to PwC’s most recent Global PPM survey, programme failures and overruns are most commonly due to poor estimates, changes in scope, and poorly-defined goals.  Projects produce and consume masses of data in various different formats throughout their lifecycles, and this is only going to increase. 

For example, the growing adoption of a digital data representation of buildings, known as Building Information Modelling (BIM) is making construction data available to programme managers in new ways.  Technologies associated with the Internet of Things can provide even more data for projects. 

We can foresee, for example, a time in the near future when the components of a construction project have the capacity to update a project plan automatically when their installation is complete. Infrastructure projects are also seeing an increase in the use of mobile applications to provide real-time updates from the field, such as sending photos of defective parts to engineers who can remotely resolve problems. 

In IT projects, the growing adoption of DevOps toolsets can provide project managers with more real-time data on the status of application development and deployment.

PwC’s 4th Global PPM survey further found that the most common service carried out by Project Management Offices (PMOs) was ‘status reporting to upper management’. While status reporting is invaluable to the programme manager in order to operate effectively, this leaves little time for them to apply their insight and focus on stakeholder management and change management.

Organisations are typically ineffective at capturing and storing of knowledge in a structured way and making it available to project managers at a time when it will be most useful and valuable.  What if we could get the right knowledge delivered to project managers in an intelligent way and at the right time given the programme stage and context? 

AI methods can be deployed to effectively use the information gathered through project management processes and create additional knowledge by mining that data to identify relationships and trends.  Cutting out excessive information noise and having the right knowledge at your fingertips would enable more effective decision making.

One of the main benefits of artificial intelligence technology is the capacity to analyse large amounts of data and to take actions based on patterns of data matched to historical precedent.  Imagine an organisation with access to a database of project plans, risks, issues, benefits measures, key performance indicators, and outcomes.  With AI-enabled project management tools, this organisation could, amongst other things, do:

1.       Reference class planning:  creating detailed programme plans and budgets based on a database of plans, using relevant tags such as location, materials, business processes, etc.

2.      Benefits benchmarking and tracking:  developing benefits measures and targets based on historical information, again using relevant tags such as industry, geography, process maturity, etc.

3.      Updating a master project plan with inputs from multiple sources:  sorting out interdependent changes and flagging changes that have undesirable consequences.

4.      Generative risk and issue management:  the AI identifies possible risks & issues based on a database of risks & issues from past programmes that are similar to the current programme; it also suggests mitigating actions based on the history of the outcomes of alternate solutions; furthermore, natural language processing could scan project emails and documents for indications of risks and issues.

5.      Generative status reporting:  the AI analyses project-to-date data (e.g. progress, spend, forecasts, risks, issues) and identifies trends, patterns, and exceptions that are brought to the project managers’ attention.

As the database grows, the AI learns new patterns.  

How do we get from where we are now to this vision where our data works for us, through the application of technologies such as AI?  We need to lay the foundations now by thinking about our project information management strategies to ensure that we are beginning to capture the right data in the right formats in a responsible way, and storing this in such a way that we can apply new tools and techniques to release its true value. 

We need to consider the systems and tools we are using to manage our projects and programmes and assure ourselves that the data captured is usable in this new context.  We need to review our adoption of open data standards and techniques. 

As a project management community, we need to start talking with project management software vendors about adding AI to their tools.  These are just the first steps in beginning the AI revolution in project management.

Latest Book

Cover for Management of Portfolios