What are the AI project management tools that are defining the next generation of project portfolio management? We look at all the different opportunities presented by AI.
This article in brief:
AI, once a remote concept from ‘the future’, seems to be taking over the world. Today, we use AI draft emails, design brochures, create schedules, and research topics. Closer to home, we use it in smart devices, personal assistants, and talking with stores online. The AI future is here right now, and it’s infiltrating many more aspects of our lives.
In project management, leaders must keep abreast of all the changes that can impact delivery. The role requires continual monitoring, calculations, and adjustments around processes to secure the best outcomes. And AI offers many helpful tools to assist.
On top of that, customers and other project stakeholders have greater expectations of what is possible from new projects. AI is an asset, not only for the automation of repetitive tasks, but for driving enhancements in new product capabilities. AI tools help guide new project decision-making and are helping to build products and services once only dreamed of.
Right now, we can get ahead of the game by understanding AI’s capabilities and how it will change the ways we work.
In this article we look at how AI is already impacting delivery processes. In time, automations and AI assisted processes will become an essential feature of project management. We explore the skills that will help you embrace future developments in AI and get you ready to leverage them.
Let’s start with the basics: what is AI? According to Gartner, Artificial Intelligence (AI) is:
“advanced analysis and logic-based techniques, including machine learning, to interpret events, support and automate decisions, and take actions.”
The key thing to understand is that AI uses strict logic and rules-based frameworks designed by programmers. In other words, its effectiveness is highly reliant on human input. We are the ones who proscribe, limit, and extend its capabilities.
In project management, experts predict that AI will manage 80% of current project management responsibilities by 2030. Given the weight of responsibility currently carried by project managers, that might seem unlikely. However, when you consider the various types of AI and how they might be developed, such predictions are not so far-fetched.
These types can be generally grouped into either ‘Machine Learning’ or ‘Knowledge-Based’ AI models.
Machine learning analyses data and builds models by detecting patterns. AI’s ability to detect patterns in raw data and apply those patterns in decision-making is hard to ignore. The key benefit is the elimination of human error and bias, improved decision-making capabilities, and a reduction in manual tasks. Machine learning acquires knowledge quite literally by observing data without prior learning (‘domain knowledge’).
Deep learning is a sub-category of Machine learning. Deep learning predicts outcomes based on probabilities. It does this via interconnected nodes which process input signals and generate results through ‘neural networks’. These networks are inspired by the human brain. Deep learning AI can independently make predictions of future outcomes and take appropriate actions. Using these functions, deep learning will interpret and process enormous amounts of data quickly and efficiently.
Knowledge-based systems, unlike machine learning, rely on the software being given pre-existing knowledge systems to make decisions and solve problems. This knowledge is typically based on rules, facts, heuristics, ontologies, and other structured data formats – like languages. Such systems are constructed to mimic human intelligence, skills, or behaviour. So, knowledge-based AI typically consists of the knowledge base, an inference engine for reasoning, and a user interface for interaction. Ongoing maintenance and improvement are based on user feedback and evolving domain knowledge.
Decision management AI creates processes based on internal rules and logic. It works in conjunction with other technologies by responding to triggers in the information it receives. Thus, Decision Management AI analyses data, assesses options, and provides recommendations based on its internal logic. Functions include data analysis, the modelling of historical data, predictive analytics, optimisations of variables, personalised recommendations, risk assessments, process automation, and even fraud detection.
Robotic Process Automations (RPA) are technologies that deploy ‘bots’ to complete tasks by following a defined set of instructions. While RPA is a form of automation, it is not inherently AI-driven. The tasks RPA completes are often mundane, repetitive tasks like data entry, extraction, and transfer, navigating user interfaces, completing forms, and extracting information from documents. RPA technologies are currently growing at an annual rate of 23% a year. In addition, analysts predict 90% of RPA vendors will offer generative-AI automation by 2025.
Data is the basis on which AI functions. Therefore it requires the most accurate and detailed data available. Without it, AI can be useless, or even harmful, to your objectives.
There are three distinct types of data used by AI:
Chat data allows AI to understand human language and queries. In fact, the main goal of chat data is interaction with humans. Today, AI has learnt to understand and construct sentences independently with a deeper understanding of human language. Thus, the reason ChatGPT is so revolutionary is because it understands intent and makes intelligent decisions around the response.
‘Our’ data is the data that is captured through your activities. In project management, therefore, Our Data includes things like:
This data is crucial in project management, but it has a specific application in AI. By applying historical data AI can optimise processes for better performance.
Knowledge data, is all the information AI has acquired by itself. This is the information it has learned from the internet or from other external sources of information. However, given that AI does not have the experience to decipher which of its acquired data is the most effective or truthful, it must rely on human experience to determine authority. That means the answers AI offers will be based on the most popular or common practices, and not necessarily the most correct.
However, to realise the efficiencies of AI, enterprises must first ensure their data management is comprehensive and reliable. Firstly, a centralised database for all project data is essential. This is because AI is only as effective as the data it has access to. So, whatever enterprise resource platform and software solutions are used to manage activities, all sources of data need to be linked to a ‘single source of truth’. Connecting data on a single platform creates significant savings in time and resource usage. Enterprise updates and reporting are instantaneous, and AI tools extend the efficiencies offered by PPM software solutions. The alternative is ‘scattered data’, where project information is kept on remote servers and in disparate locations. This kind of data management reduces productivity and creates delays in reporting.
Many project management solutions are already integrating AI. Through these features, project managers have been able to streamline their processes, and enhance the effectiveness of their PMO. There are many organisations already using AI in project management solutions for scheduling and many other PMO tasks. The convenience and greater efficiency of such tools will only increase their use over time.
pmo365 is a real-time, comprehensive PPM solution that is starting to incorporate AI tools into its offering. With ChatGPT, pmo365 is revolutionising its project scheduling, risk analysis, and reporting tools:
To create a pmo365 schedule with ChatGPT, users create a project (in pmo365’s Project app or any connected application) with start and finish dates. By clicking ‘apply template’, the user can then select the ‘start from AI’ option. From here, they describe the initiative or project and ChatGPT’s AI will instantly produce a schedule for them. It can create GANTT charts with all phases and stages articulated according to the dates inputted. The more specific the user-supplied information, the more detailed and tailored theschedule will be.
In risk analysis, ChatGPT in pmo365 lets the user ask a question around the likely risks for their project. For example, when asking what the top 10 risks for an ERP implementation might be, the solution responds with possibilities ranging from legal issues, incompatibilities with existing I.T. infrastructure, lack of employee involvement, and vendor unreliability. However, in this instance, ChatGPT was using purely knowledge-based data to generate the answers. By inputting the organisation’s own data it would generate a far more accurate and useful set of potential problems. Thus, with access to historical data on project scheduling, indexing, lessons learnt, risk assessments, and trend analysis, ChatGPT is empowered to provide stunningly accurate responses based on lived experience.
Generating reports through AI is a game changer for many project managers. Reporting is notoriously time consuming given the need to track down and reconcile any number of project files. Consider the work involved if you were asked to complete a report which listed your organisation’s top 10 projects by cost, and to create a pie graph representing this data. By the time you have sorted through all your projects, calculated their expected spend, and inputted this information into a pie graph, the actual report could be a week out of date. Even when using powerful business intelligence tools like Microsoft BI, you’d have to generate many queries to get to the answers you’re looking for.
AI removes all those steps. By linking AI to your real-time data, the user simply queries the AI Agent to “list the top 10 projects by costs and input this data into a pie chart.” Within minutes, it will have completed the task.
If AI will indeed eliminate 80% of current tasks associated with project management, the key to staying relevant is understanding what it can already do. The best way to learn what’s possible is by integrating AI into your project practices now.
Greater automation of tasks naturally changes the role of the project manager. Project managers will, therefore, be those people who can integrate useful AI tools into project practices. Project managers themselves are likely to transition to ‘soft skills’, such as ideation, communication, and problem-solving rather than the more routine tasks currently associated with the role.
Let’s delve into some of the challenges and opportunities presented by AI project management tools.
However, all these developments suggest a much greater role for the Project Management Office in the typical enterprise. The PMO, already growing in importance, will enhance its role as a driver of strategic growth, with the increasing impact of AI on project management tasks. Future PMO leaders will have to develop unique skillsets to manage complex portfolios and pursue greater returns on investment.
As one of the oldest industries in the world – and perhaps one of the most profitable – construction has been slower to adopt new tech in project delivery. In many cases it is suffering from ineffective project management, inadequate skills, and poor design processes. In part, this is due to the fragmented nature of construction, its dependency on public demand, and its informality. However, streamlined processes and AI could solve many of these issues in a very short space of time.
According to McKinsey, if the construction industry could catch up to the 2.8% global labour-productivity growth, it could add nearly $1.6 trillion to its bottom line. That figure alone is reason enough to investigate the potential of software solutions for construction projects.&nbs