(photo by Peter Shanks)
Some people have difficulty understanding the term ‘artificial’ intelligence. A simple explanation is that it is the ability of a computer program to discern patterns in data and make correlations that are not possible with a human brain. Think of a database with over a million pieces of data divided into a variety of categories. It is impossible for a human brain alone to discover similarities or dependencies in the data. A machine learning program using different classifiers is capable of doing this and is proving extremely valuable with tasks such as the interpretation and diagnosis of MRI brain scans or lung X-rays in children because the level of accuracy is much higher than that of a skilled technician.
This concept can also be applied to project management. Specifically, how does a project manager accurately diagnose a serious project issue and make the best decision to resolve the issue? An optimal solution involves three elements: the past, the present and the future.
1) The past. AI programs rely on machine learning to use historical information for training and then make a prediction for a new set of data. For project management, the software program will look for similarities based on historical information from previous projects such as lessons learned and an issue log with issues, the decision and resulting outcome. What has occurred and was there a successful resolution? This is the traditional area of machine learning which uses historical information to discover patterns.
2) The present. An AI tool for project management needs to also include the current project status and project environment. These can be important factors to determine the best solution to an issue or problem. PMBOK® defines 33 documents that are used to capture the project events although not all projects maintain these documents. Large projects typically have a project schedule and capture basic information such as budget, schedule, change orders, and EVM (Earned Value Management) metrics. The project environment may include internal factors such and the organization’s policies and procedures as well as external factors such as interest rates for loans, GDP, contractor prices, inflation, and resource availability. Factors in the current project environment may be required in order to assess the success rate of various solutions or actions derived by an AI tool.
3) The future. The final category extends the ability of an artificial intelligence program into a different area which is metrics about the future. What is likely to happen in the next few years or for the remaining duration of the project? This may seem like pure speculation and part of that is true. However, both the internal and external environments exhibit trends that can be captured and used to assist in finding the best decision for project issues. Think of interest rates. In a climate where interest rates are low and inflation is beginning to climb, a rise in rates in the future can be expected. Of course, the Federal Reserve in the US and the Bank of Canada in Canada also issue statements that they expect to raise or lower rates in the future so there is not much speculation at all. Trends such as the expected growth rates and resource availability are widely available in publications such as the Economist or published by a government organization. A project’s internal environment may reveal a trend due to the inability to achieve budget targets and result in a forecast of EAC (Estimate at Completion) based on the CPI (Cost Performance Index) run rate. While some results such as a sporting event may be unpredictable there are many future metrics that will easily fall within a limited set of parameters. These can be similar to predicting the path of an iceberg. It may move slightly in one direction or another but the direction is very clear.
The purpose of gathering all of these factors is to enhance an AI tool’s ability to consider data and make accurate assessments and predictions for a project decision. A machine learning program that considers the past, present and future will be a powerful tool to determine the best outcome in order to solve project issues.
Paul Boudreau, BA, MBA, PMP is the President of Stonemeadow Consulting and is currently involved in research about how AI technology can provide value to the project management methodology used by organizations. For more information please contact email@example.com