Short answer
Incorporate a continuous 'Operation and Maintenance' phase into AI-driven design projects and proactively map required skills to project tasks to ensure successful long-term AI solution deployment.
- Field
- Innovation & Design
- Source
- Production Planning & Control (2023)
- Method
- Action Research / Case Study
- Evidence
- Strong effect
Extending the CRISP-DM methodology with an 'Operation and Maintenance' phase and a task-skill framework streamlines AI implementation and management in manufacturing. This innovation & design research insight is drawn from a 2023 study published in Production Planning & Control. Using Action research / case study, researchers explored how this design variable affects real-world outcomes. The key design takeaway: Incorporate a continuous 'Operation and Maintenance' phase into AI-driven design projects and proactively map required skills to project tasks to ensure successful long-term AI solution deployment.
Enhanced CRISP-DM methodology accelerates AI adoption in manufacturing
Extending the CRISP-DM methodology with an 'Operation and Maintenance' phase and a task-skill framework streamlines AI implementation and management in manufacturing.
Production Planning & Control · 2023
Key Findings
- 01Significant trade-offs and hidden costs are associated with operating and maintaining AI solutions.
- 02Managing AI drift is a critical challenge.
- 03Ensuring domain, data science, and data engineering competence throughout the AI lifecycle is essential.
- 04Data engineering is a crucial but often overlooked component of the AI workflow.
- 05The trajectory of involvement for different competences changes across AI project phases.
Application
Design takeaway
Incorporate a continuous 'Operation and Maintenance' phase into AI-driven design projects and proactively map required skills to project tasks to ensure successful long-term AI solution deployment.
How to apply
When initiating an AI-driven design project, use the enhanced CRISP-DM framework to structure the project plan, ensuring that post-deployment activities and the necessary skill sets are considered from the beginning.
Project actions
- 01When designing an AI-powered product or system, think beyond just the initial creation; plan for how it will be maintained and updated.
- 02Identify the specific skills needed for each part of your AI project (e.g., data collection, model building, deployment, monitoring) and ensure you have access to those skills.
Method & Evidence
Variables
Strengths & Limitations
Strengths
- +Longitudinal study over six years provides depth and practical insights.
- +Focus on a widely used methodology (CRISP-DM) makes findings broadly applicable.
Limitations
The specific skills required and the challenges of AI drift can be highly dependent on the type of AI being used and the specific manufacturing process.
Reliability & validity
The validity is strengthened by the six-year action research approach, implying practical relevance. Reliability might be a concern if the findings are highly context-specific to the participating firms.
Think critically
How might the 'hidden costs' of AI operation and maintenance differ across various manufacturing sectors (e.g., automotive vs. food processing)?
Design Principles
"AI solutions require a holistic lifecycle approach that includes robust operational and maintenance planning, supported by clearly defined skill requirements."
This research addresses the practical challenges of deploying and sustaining AI solutions in manufacturing. By providing a more robust lifecycle framework, it helps design teams anticipate and mitigate common pitfalls, leading to more successful and valuable AI integrations.
What This Means for Your Design
Making AI work in factories is hard! This research suggests a better way to plan and manage AI projects by adding steps for keeping the AI running smoothly after it's built and making sure the right people with the right skills are involved at every stage.
How to use in your project
- 1.Reference this research when discussing the methodology for developing AI-driven solutions, particularly when justifying the inclusion of operational and maintenance phases or the need for specific domain and data science expertise.
Add to My Project
Quick Cite
Paragraph starter
The development and implementation of AI-driven solutions in design practice necessitate a comprehensive lifecycle approach. Research by Bokrantz, Subramaniyan, and Skoogh (2023) highlights the limitations of traditional methodologies by proposing an enhanced CRISP-DM framework that explicitly incorporates 'Operation and Maintenance' and a task-skill mapping system. This approach is vital for managing the complexities of AI drift and ensuring the sustained value of AI integrations, emphasizing the critical roles of domain, data science, and data engineering competences throughout the project.
Source
Production Planning & Control
Realising the promises of artificial intelligence in manufacturing by enhancing CRISP-DM
journal · 2023
View sourceQuestions About This Research
- What does the research say about enhanced crisp-dm methodology accelerates ai adoption in manufacturing?
- Incorporate a continuous 'Operation and Maintenance' phase into AI-driven design projects and proactively map required skills to project tasks to ensure successful long-term AI solution deployment. Evidence: Production Planning & Control (2023).
- Why does "Enhanced CRISP-DM methodology accelerates AI adoption in manufacturing" matter for design?
- This research addresses the practical challenges of deploying and sustaining AI solutions in manufacturing. By providing a more robust lifecycle framework, it helps design teams anticipate and mitigate common pitfalls, leading to more successful and valuable AI integrations.
- How can designers apply this research?
- Incorporate a continuous 'Operation and Maintenance' phase into AI-driven design projects and proactively map required skills to project tasks to ensure successful long-term AI solution deployment.
- What were the main findings?
- Significant trade-offs and hidden costs are associated with operating and maintaining AI solutions.. Managing AI drift is a critical challenge.. Ensuring domain, data science, and data engineering competence throughout the AI lifecycle is essential.. Data engineering is a crucial but often overlooked component of the AI workflow.
- What research method was used?
- Action Research / Case Study.
- How strong is the evidence?
- Evidence strength is rated Strong effect, based on a 2023 journal from Production Planning & Control.
- What should I do differently in my next project?
- When initiating an AI-driven design project, use the enhanced CRISP-DM framework to structure the project plan, ensuring that post-deployment activities and the necessary skill sets are considered from the beginning.
- What are the limitations?
- The enhanced methodology's effectiveness may vary depending on the specific manufacturing context and the maturity of AI adoption within an organization.