with Jane Reilly
“Using Artificial Intelligence in Project Management”
Jane has over a decade of experience within the general area of artificial intelligence and machine learning, in both academia and industry, designing, implementing, and managing research projects. Her research has been published and presented across the globe at academic conferences.
Subscribe to Project Management Paradise via one of the links above and you’ll automatically receive new episodes directly to your device.
Watch “Using Artificial Intelligence in Project Management” here:
Key points from Episode 123 “Using Artificial Intelligence in Project Management” with Jane Reilly
Could you tell us how you got interested and involved in the world of artificial intelligence?
I guess from a young age, I was always really interested in data and finding patterns, connections, and correlations between things. When I finished my primary degree, then I worked in the industry for a little while, and then I went back into academia and I got a fantastic opportunity to join a research group in NUI Maynooth University.
We were pioneering some really innovative technologies and implementations of artificial intelligence and machine learning in the area of computation and facial expression analysis. What we were looking at there was classifying facial expressions in terms of emotions. So –identifying was someone happy, were they sad, were they angry, and then how happy were they or how sad were they.
And the bit that I was really focusing on there, clearly the most important part was, were they really happy, or were they pretending or displaying their true emotion. And then I, kind of, looked into blended expressions and facial expression analysis and dynamics, and I kind of went on a really interesting journey there with that and it brought me all over the world, presenting at various conferences and meeting some fantastic minds and researchers in the area of computer vision and machine learning.
Since leaving that position, I kind of kept my toes in the water and tried to keep up to date, about the trends and what was going on, being, you know, an industry sponsor, with an interest in direct collaboration, and then, I guess, more recently, I would have been leading some Enterprise Ireland funded research grants.
And all of that brings me to the current day, working at Cora Systems as its research and development manager. And what we are looking at in our team is how we could look at integrating AI into the every day. And I guess the question we’re trying to answer is what makes a project manager great? Some project managers, kind of, have the Midas touch, for every project they work on, it succeeds and we really want to find out why. Project management is part art and part science and we’re trying to quantify what it is they do so we can guess and replicate repeated success for all of our project managers.
How do you envision harvesting this data to make a good project manager great?
We have an awful lot of project managers, an awful lot. We have a huge number of project managers who use Cora PPM every day and everything that they do we record so we have an extensive audit, part for governance and then so they can see what happens. And what we can do is we can mine that data and add some information to it so they can find out why they did what they did. And then from that we can build on some data analytics and add in some machine learning and eventually what will be hoping to do is drive some wisdom that we could then apply to all of our projects or make available to all of our project managers. We also of course respect full GDP protocols. We look at trends and patterns, again. None of this would be specific to any particular clients. It would be the general trends.
Can you tell us, why should we not fear AI?
I guess when you think about AI and you can look back at the movies, you would have watched as a child or any other age, such as the Terminator, or kind of think like some of those films would have said: “In 2020, we will all be driving these cars that would be elevated and there will be no people driving. It will all be completely machines.” And it can be a bit scary. And you can’t really look at AI without getting into ethical issues, but there are levels to AI and what we’re looking at in Cora Systems is really augmenting the project manager’s experience.
So, it’s almost like a virtual assistant. So, we’re helping or providing possible solutions. And then the human or the project manager will affect, will at the end decide. They can ignore, they don’t need to follow through on that one. And then we use Alexa every day. That’s AI. We use Google Maps and we’ve got our Spell Checkers, you know, it’s in our lives already at certain levels and it really will be that non-intrusive level that we’re targeting in Cora Systems. If you have those AI elements in online shopping and you have the “I’ve noticed you’ve got this in your cart. Would you also like this kind of thing?” So, we say “Okay, based on where your project is right now, these kinds of things might make it more successful.”
So, what foundation does Cora PPM stand on to help support these intelligent, informed, decisions for our managers?
So where we’re at now in Cora Systems, we have enough large data and some of our project managers, they would be running reports on particular days during the week. So a classic one, I guess, would be additional benefits from AI here would have that report ready. So that they can look at it. They don’t have to add anything. It’s a kind of a thing that it’s ready for them to review or amend, so we could skip a few steps, to make it easier for them. And I think we could do it around a delegation of work. So if someone is on leave, to automatically delegate to someone or if there is a time-sensitive piece of work needs to be done, to remind people. We are really just prompting people to take action when needed. So, it’s really just helping, reminding, refining.
Tell us a little about how artificial intelligence can help ensure success for Project Managers in uncertain situations like, for example, Covid-19 in 2020?
I think you can’t look at anything in 2020 without putting it in the context of Covid-19. And then in lots of ways, Covid-19 can be considered a little bit of a Black Swan in that it’s some unprecedented event that occurred. We had no way of knowing it was going to happen and some of our companies would have put a pause on their operation. So, in March, they were operating fine, and then suddenly they paused and they lost up to two quarters. So their forecasts would be completely out. However, some of our clients have weathered the storm quite well, they’re able to adapt, adjust, amend and they’re still doing okay. So, what we could do there like we said earlier about the patterns. What is it that made those projects so successful? Are there key indicators? Are there things that everyone should watch that maybe we all aren’t?
So we can learn from the collective experience and I guess suggest to project managers what they may change or what they could change. When we think about this, we’re trying to quantify the magic. So, some people appear to have a natural ability. Their projects are always successful and why is this? It’s not somebody’s intuition. Some of it is scaled. What parts are joining our systems? And how can we ensure and make that information available so that we can turn good project managers into great ones?
How quickly is the world of artificial intelligence moving?
When I started on AI, we were really hampered by the hardware. So, it took a long time for the models to learn, you could be talking about days and weeks and then you would have missed something, so you had to go back and adjust, amend, refine and go again. Because our processes are so powerful now, we can get results really fast. When I started, it took a long time to process information.
When I was studying facial expressions, I had to, at the beginning, individually annotate the entire video to catch every single blink, wink, nod, move, you can imagine. So, it took a while. Then we developed software to track those features and you to train them as well, it took a lot of time in the day to prep and add the feature extraction. Now that processes are faster. I guess the hardware is cheaper. We can harness the cloud, so we can get the turnaround a lot faster. The biggest blocker is trust, so getting our projects and portfolio managers to trust us, that we are trying to make their jobs easier not obsolete.