In this blog, we introduce the concept of automated project management and its present state of the art.
The phrase automated project management (APM) is at least 40 years old, the concept was first mentioned by Smith and Mills in 1983. The success of AI in recent years and the availability of extensive dataset has rekindled the interest in AI-based project management.
Starting from the humble “what-if” rules in the 1980s, the APM today can do much more than just generating automated reports.
Before dwelling on how AI can be used in Project management (PM) let us first understand:
Planning, Scheduling, and controlling project activities in order to optimize performance, cost, and time goals.
Artificial Intelligence deals with making agents that can learn from their environment and experience. Both AI and PM are multi-layered fields, covering diverse areas and terms. Conventionally, a project life cycle consists of five stages: initiation, planning, execution, control, and finally closing.
In modern agile project management, the stages are approached iteratively. Teams using agile methodology can adopt AI-based bots that can assist in managing tasks, providing daily updates, and even alerting if any issues arise.
That is not all, the graphic below lists some PM areas where we can further leverage AI.
One of the most crucial, but error-prone stages in the project life cycle is at its beginning - the stage of requirement analysis. According to James Martin, more than half of all the engineering errors originate in requirements. Thus, it is important to find and correct the requirement errors at the very beginning of the project itself.
At present, the requirement analysis is done by humans and involves one or both of the two techniques checklist and peer review. Both the processes are error-prone and time-consuming, especially when dealing with large documents. The recent developments in the field of Natural Language Processing can be leveraged to develop an automated requirement analysis approach and hence reduce the errors.
The historical data, from previous projects, can be used not only to design a data-driven decision policy but can also be used to predict time and cost goals. This will increase the speed of decision-making. With AI agents monitoring and reporting task status, any bottlenecks can be quickly recognized and dealt with.
AI can be used to schedule and allocate resources, including the most important resource, the human resource. Employing past data of employees, their qualification and technical skill, as well as their social network and success rate when they work together, AI models can be trained to build a strong and successful team.
There exist a large number of AI algorithms like SVMs and regression analysis which can be trained to perform complexity and risk analysis. And in the end, the project solution can be evaluated in terms of success and meeting goals.
The scope of AI in project management is vast, it is limited only by your imagination and the availability of historical data.
Smith, L. A., & Mills, J. (1983). Reporting characteristics of automated project-management systems. International Journal of Project Management, 1(3), 155-159.
Russell, S. J., & Norvig, P. (2010). Artificial intelligence – A modern approach (3rd ed.). Upper Saddle River, NJ: Prentice Hall.
Schoen, M. (2017). Hype cycle for project and portfolio management, 2017. Retrieved from https://www.gartner.com/doc/3772090/hype-cycle-project-portfolio-management
Martin, James, An Information Systems Manifesto, Prentice-Hall, January 1984.
About the author
Amita Kapoor is an Associate Professor in the Department of Electronics, SRCASW, University of Delhi, and has been actively teaching neural networks and artificial intelligence for the last 20 years. Currently, Amita is Head of Data at digitty.io.