Choosing the right production line as a first step to AI deployment in manufacturing is critical since it will influence the organization’s progression towards successful AI integration. Here’s how to make the right choice…
—
Leading manufacturers are using AI to solve perennial problems such as unstable or low yield and throughput, quality issues, CO2 emission levels, and waste. In this way, AI helps manufacturers produce better products at a lower cost, and as a result, markets are becoming more competitive.
Because of this, more manufacturers are looking to AI technologies as a way to improve their operations’ performance.
The AI deployment process, however, can seem like a considerable challenge for many manufacturers.
A common question is how can manufacturers deploy AI into their existing operations.
Recap: The Lighthouse Strategy
In a previous blog post, we outlined the proven strategy used by leading manufacturers to successfully deploy AI:
Focus on a specific production line to begin with, and then scale to other lines from there.
Known as “The Lighthouse Strategy”, this approach is effective because it enables manufacturers to build a clear “proof of concept” that can demonstrate value relatively quickly.
Additionally, any failures that do occur are manageable since they are controlled within the boundary of a single line, and can be quickly resolved and learned from.
So, how do manufacturers identify which production line to start with?
How to choose the right production line to start AI deployment
For many organizations, their first deployment process will determine the future success of AI implementation in the company.
For this reason, it’s critical for manufacturers to choose the right production line to begin with since a positive result will draw support from teams and justify budget allocation moving forward.
Here are 4 things for manufacturers to consider when choosing the right production line for their first AI deployment:
1. Capacity constraints
Look for a production line that has a high demand from the market. The performance level of this line should have an effect on sales, and therefore plays a significant part in the overall business. Deploying AI to this line will enable you to gauge the business benefits brought on by AI deployment.
2. Losses and inefficiencies at the line
By comparing the level of losses on the lines, we can see how much potential gain we have by optimizing the process with artificial intelligence. A production line with very apparent production losses (throughput, quality, waste, etc.), or high potential for improvement in other key areas (e.g. emissions), is the right choice for AI deployment since there is great opportunity for improvement.
3. Data maturity of the line
What level of data is the production line already producing?
How far along is this production line to being able to provide good data for analysis?
By answering these 2 questions, we can compare how ready each production line is for providing the type of data needed for AI to make significant improvements.
Question: Does this mean the production line producing the most data is the right one to start with?
Answer: Not necessarily. While data maturity is a factor (as seen in point 3. above), choosing a line with capacity problems and significant losses far outweighs the amount of data produced by a line. A data-rich production line that’s performing at peak levels might yield an excellent analysis, but won’t give us the meaningful business benefits we are looking for at this stage.
Related: How Barilla went from no data to 36% waste reduction with AI – in just 4 months
Deployment concerns
Each organization is different, and so are the challenges brought about by taking on a project such as AI deployment. These may range from the performance levels of the production teams, to how the deployment process fits into the current business dynamic. The more we know about each production line, the better we can decide on the best candidate for deployment. Take the time to map out any additional concerns regarding AI deployment to each line.
The first step towards successful AI deployment starts here
Once your production lines have been analyzed according to the criteria in the 4 channels outlined above, it will be much easier to identify which line should be the first in the organization’s AI deployment process.
This will set your company in the right direction towards successful AI deployment and significant business impact early on.
Watch the full video on how to choose the right line to begin your Industrial AI adoption: