Image from ChatGPT 4o with a prompt of “draw me an AI graphic for effective use of artificial intelligence tools.” — The image above is from ChatGPT 4o with a prompt of “draw me an AI graphic for effective use of artificial intelligence tools.”

This is the last installment in a series about the effective use of AI tools in project management. In part 1, we explored the potential of AI in project management, emphasizing the importance of data quality and integration with existing workflows for successful implementation. In part 2, we dug a little deeper into AI with 3 more ways that AI is revolutionizing project management. Then we will explore 5 ways to ensure seamless integration for your organization.

Part 3 combines real-world examples and looks at Prompt Engineering with sample prompts.

Real-World Examples and Sample Prompts

Let’s explore a few examples to see how prompt engineering works.

Cost Improvement

We were working with an organization that specializes in procurement. Their customer was a private equity firm that had purchased a company, and they were trying to find a way to improve the cost models. The procurement company tried to find other cost-effective material acquisitions and used “should cost” modeling to determine if the product costs the suppliers provide are within the range of expected or possible costing. The material they wanted to re-source was aluminum, as it was consumable in manufacturing. As product development experts and system manufacturing engineers, we understand there are many ways to improve the cost structure. We are working on a short book on improving our cost structure. One of the alternatives we explored through AI was material substitution. We went through several prompts with an AI for the industry and the specific application. After some time, adding details, and refining the prompt, we found a material substitute used in that industry for that specific process. The customer preferred a quick cost improvement, and material substitution requires finding the correct material and testing to ascertain the processes associated with this change. As such, this will require time and investment. The upside would have been moving from a mined material to a created material.

Risk Identification

As a check, we used AI to provide a list of risks associated with developing embedded products with a team distributed across the globe.

Prompt

List the most common and severe risks associated with developing embedded products with a globally distributed team – label each in a category as severe, or common, or both.

Response

Common and Severe Risks

Communication Barriers: Miscommunication due to language differences, cultural misunderstandings, or time zone discrepancies can lead to misaligned project goals and expectations. This risk is both common and severe as it can significantly delay project timelines and impact product quality.
Coordination Complexity: Managing and coordinating tasks across different locations adds complexity to project management, increasing the likelihood of errors and inefficiencies. This common issue can also be severe, affecting project delivery and cost.

Common Risks

Data Security and Intellectual Property Concerns: Working across borders increases the risk of data breaches and IP theft. While common, the severity can vary based on the nature of the data and the security measures in place.
Technology Stack and Tool Inconsistency: Differences in preferred technologies, tools, or platforms among team members can lead to integration challenges. This common issue usually poses moderate risks but can be mitigated with early planning.

Severe Risks

Regulatory Compliance and Standards: Non-compliance with international and local regulations related to embedded products can lead to legal issues and product recalls. This risk is severe due to the potential legal and financial repercussions, though it might not be as common if proper due diligence is conducted.
Quality Control Variances: Ensuring consistent quality standards across different teams and locations can be challenging. This risk is severe because it directly affects the final product’s reliability and safety, although it might not occur in all projects.
Dependency on Key Personnel: Relying heavily on specific team members in different time zones or locations can severely impact project progress if those individuals become unavailable. This risk can lead to severe project delays or failure, particularly in projects with tight deadlines.

Prompt Engineering

Prompt Engineering is a new field, but it is not one in which one can get an engineering degree. It involves creating a clear, concise direction or prompt for the AI tool to respond with the information that one requests.

Writing a prompt is an experiment. One starts with a prompt based on what response is wanted.

For example, “Please provide the Work Breakdown Structure for constructing a building.”

That’s good, but what about square footage, number of floors, type of building (medical, office, warehouse, etc.), architectural style, locale, and so on?

The more information provided in the prompt, the better the response. However, one can think about the interaction with the AI tool as a conversation. Enter a prompt and get a response. Based on the first response, the next prompt can ask for more detail.

Here is an example.

Prompt: Please provide the Work Breakdown Structure for constructing a building.

Response: Creating a Work Breakdown Structure (WBS) for constructing a building is a detailed process that involves breaking down the construction project into smaller, more manageable components. The WBS is a key project deliverable that organizes the team’s work into manageable sections. Below is a simplified version of a WBS for building construction. This structure can vary based on the project’s size, complexity, and specific requirements. (Plus, a detailed list is not included here.)

Next Prompt: Great. This building will have 4 floors and be used as a medical facility. Any changes to the work breakdown structure?

Response: Constructing a medical facility, especially one with multiple floors, introduces unique requirements and considerations compared to standard building projects. The Work Breakdown Structure (WBS) for a four-floor medical facility must incorporate specific elements related to healthcare operations, patient safety, accessibility, and specialized medical equipment installation. Here’s how the WBS might be adapted (plus a detailed list of additional items not included here).

Change of Building Usage

New Prompt: Great. This building will have 2 floors and be used as a warehouse with offices on the second floor. Any changes to the work breakdown structure?

Response: For a building designed to function as a warehouse with offices on the second floor, the Work Breakdown Structure (WBS) would indeed require some modifications to accommodate the unique needs of such a hybrid structure. Here’s an adapted version of the WBS reflecting these changes (plus a detailed list of additional items not included here):

Learn More about Prompt Engineering

There are several books available on Prompt Engineering.

The Art of Prompt Engineering with ChatGPT: A Hands On Guide (Learn AI Tools the Fun Way), Nathan Hunter
The Art of Prompt Engineering with ChatGPT: Pass the Book Edition (Learn AI Tools the Fun Way!). Nathan Hunter
ChatGPT Revolution, How to make money with ChatGPT for beginners? by Charles Henry Carter
AI For Marketing and Product Innovation, Powerful New Tools for Predicting Trends, Connecting with Customers, and Closing Sales Dr. A. K. Pradeep, Andrew Appel, Stan Sthanunathan
The ChatGPT Millionaire, Making Money Online Has Never Been This Easy, Neil Dagger

Outlook and Emerging Trends:

We can foresee when these AI tools may be created for an industry or a specific company. This can help ensure the input data’s veracity and the algorithm’s suitability to support the business needs. The Gemini debacle is a stunning failure, though not specifically for business.

Simulations are not new. Rather, they are based on mathematical models. You may have used Monte Carlo simulation for schedule predictions based on variations of the composite tasks. AI can be of similar assistance. We know of a company that records the results of its development efforts in a database. It captures data on design alternatives that were explored and descriptions of the results of those explorations, specifically failure modes. All of these would be fodder for the input or source data for the evaluation.

Enhanced Decision-Making with AI

AI’s ability to analyze vast amounts of data in real time is ushering in a new age of decision-making in project management and product development. AI can identify trends, patterns, and insights that are invisible to the human eye, providing a data-driven foundation for making critical project decisions. Using this tool speeds up the decision-making process and enhances its quality, leading to better outcomes and more innovative products.

Automating Routine Tasks: A Focus on Strategic Activities

AI automates routine and time-consuming tasks, from scheduling and resource allocation to budget tracking and compliance checks. This automation frees project managers and their teams to focus on their projects’ more strategic and creative aspects. As AI handles routine tasks, project teams can dedicate more time to solving complex problems, brainstorming innovative solutions, and engaging with stakeholders, thereby adding greater value to their projects.

Collaborative Robots and AI in Product Development

In product development, AI is not just behind the scenes, but increasingly front and center in the form of collaborative robots (“cobots”). These AI-powered machines work alongside human colleagues in research labs and on production floors, enhancing precision, efficiency, and safety. The collaboration between humans and AI in product development is accelerating innovation, enabling companies to bring new products to market faster and more reliably.

The Future of Work and Project Delivery

The role of AI in shaping the future of work and project delivery methodologies is profound. Traditional, linear project management approaches give way to more agile, adaptive, and data-driven methodologies. AI is at the heart of this transformation, enabling project teams to respond more swiftly to changes, make smarter decisions, and deliver projects and products that better meet the evolving needs of customers and markets.

Summary

Presently, AI is a mixed bag. However, we should do well to recall how computers first came out. It is claimed that a CEO of Digital Equipment Corporation circa 1977 said, “There is no reason for any individual to have a computer.” It may be like AI. Like any emerging tool, there will be fits and starts. It is also true that the algorithms, evidence shows, are easily tainted with human biases. Likely, we should never consider the output of AI as the single source of truth. As humans, we should remain skeptical about whether we use AI or not.