We are establishing a new battery cell assembly line. My employer inquired what we could do with the data generated by this manufacturing line. I do not know what to say other than predictive maintenance. What else can you do with this data? What are the most common Data Science applications in manufacturing companies?
3 Likes
I work as a data scientist for a manufacturing company. All of your recommendations are pertinent. Here are some other things I work on:
1.Production line scheduling (shorten setup times, increase output in less time).
2.Optimal purchasing: Acquire the right supplies at the right moment. Prices can fluctuate a lot and have manipulable constraints depending on the kind of production organization. This can be effectively connected to demand forecasts.
Given that demand forecasting impacts practically every aspect of an organization’s operations, I believe it offers the most “bang for the buck” in most cases.
1 Like
On Discrete Event Simulation, never give up. extremely effective production tool.
1 Like
@Dario Is it correct to say that you utilize that to perform what-if analyses? That is, you simulate the production line, play around with a few factors, and observe whether the line statistics as a whole change? For example, what is the increase in throughput if this step is X% faster? Or what purpose does DES serve?
1 Like
@Royal Yes, that’s correct. We typically study the relationship between a resource and a KPI through a Monte Carlo-like analysis.
Some examples include:
- Which machinery should I purchase?
- What buffer sizes are optimal?
- How many resources are needed to avoid bottlenecks?
- Can I remove any resources without affecting throughput?