My question of the week is to have your feedback on which industry is best to start using AI in their Project Management Office.
Let me give you my opinion and an example.
AI works best with data. Lots of data. If a company has been operating in projects for more than a year and that the project duration is about 2-3 months, then I believe we have enough data to start working with. Also, if the PMO runs similar projects over and over, it really helps to find links inside their data and to teach the AI platform at the initial stage. Of course, this PMO needs to be opened to innovation and digital transformation in order to obtain the adoption as well.
For example, I believe that if a marketing firm that runs projects for 2-3 similar clients on similar type of projects for over a year, they are best positioned to start experimenting AI in Project Management.
What do you think? Saving Changes...
Sort By:
Luis BrancoCEO| Business Insight, Consultores de Gestão, LdªCarcavelos, Lisboa, Portugal
Dear Edward
Interesting your question
Thanks for sharing
There are many industries and companies where Artificial Intelligence can be applied because the amount of data that is available is huge
There are industries where artificial intelligence is already being used
Examples:
- Aircraft Construction
- Aeronautics (civil and military)
I believe that artificial intelligence is already being used in stock exchange operations Saving Changes...
if your question relates to the use of AI to support better decision making when planning and managing projects, then almost any kind of service provider should have access to a sufficiently large data sample.
Product development organizations might be harder pressed for this as the level of uniqueness of their projects would be higher.
AI requires not just data, but the right data, and access to the data. That naturally fits domains well which are primarily digital, and are networked to external sources like online gaming, and banking systems. Physical systems like manufacturing are much more difficult. Often there is still a lot of work done by humans. If you want specific data, someone has to record it which adds time, and is error prone. Automated machines collect data, but it is limited to the functionality such as the feedback and control systems. If you want more data, you need to add more sensors which again adds cost. On the business management side that there may be a lot of data out there such as manufacturing, engineering, supply chain, finance, etc., but they tend to be in isolated systems that don’t talk to each other and were never designed with that as a concern.
Another problem with AI and physical systems is that the physics can be quite complicated and difficult to model such that it supports analytics. The term “digital twin” was one of the top executive buzzwords in 2016, but few companies have pulled it off successfully with predictive capabilities. Jaguar has a digital factory although I don’t know the extent of the capabilities. Boeing has some success in the area although they just announced cancelation of a 6 year factory robotics project so they’re seeing the challenges. Autonomous transportation is doing a lot in the field but modeling complex physical systems to the point where read time data can be used to make decisions is challenging from many angles.
PM AI applications don’t necessarily have to make real time decisions like autonomous vehicles as opposed to planning decisions like organization structures that integrate efficiently, but you still need enough data to enable predictive capabilities involving many different independent variables. Saving Changes...
if your question relates to the use of AI to support better decision making when planning and managing projects, then almost any kind of service provider should have access to a sufficiently large data sample.
Product development organizations might be harder pressed for this as the level of uniqueness of their projects would be higher.
Kiron
Agreed! Saving Changes...
Sergio Luis ConteHelping to create solutions for everyone| Worldwide based OrganizationsBuenos Aires, Argentina
Key to understand is: data is waste if it is not transformed into information. To understand that the key is to understand Shannon Information Theory then to understand the whole data warehousing architecture which is critical to create, manage and transform data into information. You can argue that you can use data lakes to get the data you have today. It works. But at the end of the day remember: data is waste if you do not transform it into information. To do that, you do not need AI, you need to understand your needs for information. All these is not new. It exists from 1990 and before. I am putting this comment just in case you like to go to the sources thing I recommended. Saving Changes...
Jen Jee ChanManaging Director| DotProjects Pte LtdSingapore, Singapore
Edward,
Having been in the Corporate Real Estate/Construction Industry for close to 20+ years now, I feel there is a lot of leverage to be gained if data from completed projects is harnessed to help decision making. The industry is considered a bit of a laggard compared to its more glamorous peers in ICT, Healthcare, Hospitality and other more progressive sectors but then again, the argument is the up-side could be tremendous..
Risk-based predictive models that take into account various risk categories could be used to predict project outcomes, help in feasibility decisions and indirectly enable prioritization of resources on where to tweak project systems/processes.. we are currently working on this to try to elevate such decision making the industry..
Happy to hear thoughts and discuss further.. cheers... Saving Changes...
