Types of AI Technologies in Project Management
In our dynamic project management landscape, Artificial Intelligence (AI) is an ever-evolving subject. As a result, it can be challenging to understand which AI technologies apply to project management. In this guide, we’ll explain the five distinct types of AI technologies that project managers are using. In addition, we’ll look at how you can harness their capabilities to enhance your project management activities.
What is Artificial Intelligence?
Gartner defines Artificial Intelligence as ‘advanced analysis and logic-based techniques, including machine learning, to interpret events, support and automate decisions, and take actions.’
While all AI fall under this definition, they have different forms. A key thing to remember, however, is that AI isn’t the “all-seeing” robots that it can be characterised as. In contrast, AI only uses strict logic and rules-based frameworks, designed by programmers. These frameworks can make processing complex situations and data difficult. While AI holds untold potential, its current effectiveness relies heavily on human input, thus limiting its capabilities to the framework upon which developers have built it.
Project Management and types of AI technologies
We will see drastic change in the project management sphere with the introduction and integration of AI. Early studies suggest that AI will conduct 80% of current project management activities by 2030. These statistics have often engendered reactions of either fear or excitement for project managers. However, the difference in these reactions often depends on how well you understand the genuine capabilities of AI.
The PMI report, AI Innovators: Cracking the Code on Project Performance gives us a valuable insight into the impact of the types of AI technology on project management. We’ve distilled that report down to this list of the top five types of AI technologies that will affect project management, and how they will do that.
AI for Project Management: Top 5 types of AI technologies
1. Machine Learning
Machine Learning analyses data to build models by detecting patterns. As a result, they have an improved decision making capabilities with minimal human intervention. As project managers, we all know how much data is generated and required by projects. Without AI, sorting and applying this data involves a significant amount of human effort and judgement. As machine learning continues to understand the human decision-making process, humans will have less involvement in ‘sorting’ data in project management. This will have considerable time and money savings for organisations.
2. Decision Management
Decision Management technologies create processes based on rules and logic, which then automates decision making. Good decision-making is central to successful project management. As a result, we expect AI with this capability to become frequently used by project managers. Decision management technologies are likely to work in conjunction with other technologies by reacting to triggers in input information. For example, a connected technology will input information around sensors, events, or time into the AI. The AI will then recognise this information as a trigger, and may suggest decisions to rectify the issue. Ultimately, this significantly reduces the workload of the project manager.
3. Knowledge-based systems
Knowledge-based systems (KBS) mimic human intelligence, skills, or behavior in a particular field, topic or skill. Most current systems cannot exceed the capability of humans, with the recent exception of IBM’s Deep Blue.
In project management, Knowledge-Based Systems would likely train themselves by gathering internal data. This data could concern communications, risk management, cost management and resource management processes. From here, the KBS can gain insight into the nature of the project or projects they are analysing. This insight allows for robust analysis, reducing error rates while creating considerable savings in labour.
4. Deep Learning
Deep learning uses neural networks to build, train, and test models. From those models, deep learning AI can predict outcomes based on probabilities. These neural networks are inspired by the human brain, so naturally, their capabilities are extensive. They consist of interconnected nodes which process input signals, thus generating results through their neural networks. Developers adjust the parameters built in this type of AI technology, thereby teaching the AI about the complex patterns within data. This processes continues to improve their accuracy.
Chief data officer at PMI, Mark Broome, suggests that “deep learning models will assist in predicting work effort activities, tracking project progress, and updating forecasts as the project progresses.” By utilising the power of these neural networks, project managers can both analyse and interpret copious amounts of data. This gives project managers a platform to make informed decisions from, and adapt approaches as needed.
5. Robotic Process Automation
Robotic Process Automation (RPA) might be viewed as the most logical use of AI in project management. RPA is a business process automation that deploys metaphorical ‘bots’ to complete tasks by following a defined set of instructions. These instructions are typically limited to frameworks built on ‘if, when, or else’ statements to complete tasks.
RPA is currently already taking over many menial administrative tasks of project managers from invoice approvals to reporting. This will significantly reduce the load of project managers, thereby giving them more time to spend on value-adding activities. At its current state, RPA has already impacted 21% of organisations, and is expected to impact 62% of organisations in the future.
What does AI in project management look like?
After discussing these types of AI, it’s becoming clearer how AI will impact project management. To learn how we are implementing AI in our PPM solution, pmo365, check out our blog on using AI in Project Management, how to prepare your PMO to use AI, or talk to our PPM experts!
IIR: Introduce, Integrate, Replace
Step 01
Introduce
You cannot run a portfolio on Excel and PowerPoint alone.
Project portfolio management is the discipline of seeing every project in one place, prioritising the work that matters, allocating people against demand, and governing delivery with real numbers. It is not optional at any serious scale. The moment you have more projects than one person can hold in their head, you need a single, current view of status, schedule, cost, resource and risk.
Excel and PowerPoint feel free because there is no licence conversation. The real cost is elsewhere. It is the hours spent maintaining workbooks, the version confusion, and the numbers that go stale the moment they are pasted.
A spreadsheet cannot tell you, on demand, which projects are at risk, where your people are over-committed next quarter, or how much of the portfolio budget is actually spent.
Introducing a proper PPM platform is the first step. Not to add another tool for its own sake, but to give the portfolio one place where the data lives together and stays live.
Step 02
Integrate
The instinct after buying a PPM platform is to make everyone move into it. That is the fastest way to fail. Project managers already have tools they trust, and finance already has systems of record. Force a migration on day one and you get resistance, shadow spreadsheets, and a dataset nobody believes.
Integrate first. Meet the data where it already is. Two directions matter.
Direction 01
Enterprise systems
Connect to the finance or ERP layer so actuals, commitments and budgets flow in automatically. Reporting stops being a monthly reconciliation and becomes a live view. Nobody rekeys a spend figure again.
Direction 02
The tools PMs already use
The direction most platforms neglect, and arguably the more important. The portfolio should read from the PM's own tools, not force people to abandon them.
The reason this matters is simple. That data is already there, and it is kept current by the person closest to it. When the portfolio reads directly from these sources, the status report updates itself. No chasing, no copy and paste, no reporting lag. The PM keeps working the way they always have, and the board gets a live picture as a side effect.
Step 03
Replace
Integration buys you two things: trust, and live data. Once both are in place, you look at what can go.
Every organisation carries tools and spreadsheets that either do not do the job well or carry a heavy maintenance overhead. The classic example is the resource spreadsheet. It is a workbook someone maintains by hand to track who is on what. It is always slightly out of date, owned by one person, and impossible to reconcile against real demand.
Replace it with the equivalent function in your PPM.
A proper demand management capability does what the spreadsheet was reaching for, with none of the overhead. It models demand against capacity across the whole portfolio, updates as projects shift, and needs no manual upkeep.
Replace deliberately, one function at a time, and only after the platform has earned it. The test is simple: if a spreadsheet is high overhead or low quality, and the platform does the same job natively, retire the spreadsheet.
The payoff
You stop producing reports and start reading them
Follow IIR and the nature of reporting changes. The status view is current because it is fed by the tools people already use and the systems that already hold the money. The overhead that used to consume the last week of every month disappears, because there is nothing to assemble.
That is the whole point of real-time reporting. Not a prettier deck, but a portfolio you can look at any day of the month and trust, at a fraction of the effort it takes today.
Built on Microsoft 365. Native ground for IIR.
pmo365 integrates with the tools your teams already run in, so the path from Introduce to Integrate to Replace is a natural progression rather than a rip and replace.