“Tool Life monitoring is important by having a proactive approach which involves tracking tool condition in real time, understanding wear patterns, and maintaining the tool accordingly”, says Kavita Kaushik, Head Quality and Six Sigma India Region, Cummins India.
By Neha Basudkar Ghate
Q- How do you see technologies like AI, ML, and IoT integrating with Six Sigma methodologies in today’s manufacturing landscape?
When you really look the integration of AI, ML, and IoT enhancing the traditional Six Sigma especially its methodologies like DMAIC or Design for Six Sigma (DFSS/DMADV)—they complement each other quite well. Six Sigma can be applied when you want to improve an existing process (DMAIC), or when you’re building or designing a new product (DFSS/DMADV).
But to use these methodologies effectively, what you really need is a good amount of reliable data either to reduce variation or to analyze historical performance so you can design better solutions. This is exactly where IoT, AI, and ML come into play. They help you gather that dependable data.
With AI/ML agents and tools like Power BI layered on top, a lot of meaningful decision-making can happen. You can even perform root cause analysis much before diving into process improvements. In this way, I strongly believe that the integration of AI, ML, and Six Sigma methodologies go very much hand in hand.
Q: Can you share a real-world example of how AI, ML, and IoT are transforming data handling and decision-making in manufacturing?
India is adopting the automotive model where you churn out high volumes, and today, IoT is present everywhere. Take, for example, a manufacturing line at Cummins, where we produce engines. There are so many parameters of an engine that are extremely critical, whether critical to quality or to the customer.
You cannot expect a human to manually collect and monitor all this data. That’s where IoT sensors, AI, ML, and other intelligent systems come in. These technologies continuously capture data, be it from nut runners or various engine sensors giving you critical, real-time information about the engine’s performance. This data is massive, and earlier, a person had to download and synthesize it before any process improvements could be made.
Now, AI and ML take care of that heavy lifting. They not only process the data but also provide insights like identifying trends, flagging deviations, or confirming that everything is running smoothly. So the time previously spent by a person in just gathering and analyzing data is now saved.
However, it’s important to understand that while AI can synthesize data, the decision-making still lies with the human. The human brain will always be the most important part of the equation. It’s crucial to use this data intelligently. AI agents can do a lot of processing, but knowing how to use that processed data for meaningful action, that’s where human judgment comes in.
Q: What role does precision machining play in engine manufacturing, and how do Six Sigma principles support this process?
When you really look at engine manufacturing, a lot of the components especially fuel system parts are precision machined. These are very small parts, but they come with tolerances measured in microns. If those tolerances are not maintained, the engine may fail. That’s how critical precision machining is to engine performance.
Even something as small as a collet, if not properly formed or machined, can lead to engine failure. So, precision machining becomes the backbone of manufacturing, especially in the engine space. It’s all about getting it right the first time, starting from the tool-making stage. Having the right tools, using the correct tool materials, and being able to monitor the tool’s life cycle are all essential factors.
In precision machining, most processes revolve around CP and CPK process capability indices. These are crucial because Six Sigma, at its core, is about reducing variation. CP and CPK help you understand whether your process is centered and how wide the process spread is.
If you’re making 10 million parts, the aim is to have the lowest possible number of defects. While achieving zero parts per million (PPM) is ideal, it’s nearly impossible so the goal becomes limiting it to 3.4 PPM defects, as defined by Six Sigma standards. That’s why Six Sigma is a natural partner to precision machining.
And it’s especially important because these precision-machined parts aren’t cheap. If something goes wrong like if a parameter is off, you often can’t rework the part; it has to be scrapped. So the stakes are high, and maintaining consistency and control through Six Sigma becomes essential.
Q: According to you, how important is real-time monitoring of tuning parameters and preventing premature tool wear and failure when it comes to generators, power system engines, drive trains, and the wider portfolio?
Let me take the example of Cummins. At Cummins, we have a large machine shop where we machine components like crankshafts, camshafts, and connecting rods. These parts have extremely tight tolerances often in the micron range and if those aren’t maintained properly, it can be catastrophic for the engine.
To begin with, it’s critical to use tools of the right quality. But beyond just having good tools, maintaining them becomes a shared responsibility especially for the operator on the shop floor. The operator needs to ensure the tool is functioning properly, right from the morning setups, first-piece inspections, and other process checks. All of these steps are crucial and can be managed proactively.
That said, the most important factor, in my opinion, is tool life monitoring. In precision machining, if a tool doesn’t perform as expected like if it wears out prematurely or becomes blunt and if this wear goes unnoticed, the resulting parts can have defects. You might only discover it much later, when the component’s parameters are already off-spec. By that time, it could be too late which leads to rework or, worse, scrap.
Tool wear is often invisible until it’s too late. That’s why proactively monitoring tool parameters is extremely important. There are two approaches including, reactive and proactive. The proactive approach involves tracking tool condition in real time, understanding wear patterns, and maintaining the tool accordingly.
But if failures still occur, it becomes about more than just correction, it becomes about prevention. So the question becomes: how do you prevent this from happening again? That’s where practices like intelligent tool monitoring systems, poison tests, and predictive maintenance come in. These strategies help ensure that such failures don’t repeat in the future.
