Knowhy Industry Insight

World Robot Conference × Process Agent

The World Robot Conference has just concluded.

For five days, the industry was captivated by robots playing table tennis, collaborating with humans, and performing increasingly sophisticated tasks.

“Human–machine symbiosis” was everywhere.

But walk back into your machining shop.

Stand beside your most expensive five-axis machine.

When the next job arrives, ask a simple question:

Which tool should we use? What cutting parameters should we set?

In many factories, the answer still comes from an experienced machinist.

So we at Knowhy want to ask:

Everyone is celebrating the “robotic arm’s body.”
But who is managing the “process brain”?

01 | Everyone Is Talking About Symbiosis. But the Shopfloor Is Still Stuck on Process Know-How.

Much of the “symbiosis” discussed at robotics events is physical:

Machines lift what people used to lift.
Machines move what people used to move.
Machines inspect what people used to inspect.

These tasks are standardized and repeatable.

Making humans and machines work together on them is relatively straightforward.

The harder challenge is something else:

Every cutting operation is a unique process problem.

Different material grades.
Different machine conditions.
Different tool wear.
Different geometries and tolerances.

The answer changes accordingly.

If that process knowledge is still locked inside the head of an experienced machinist, then no matter how many robotic arms you deploy, you may simply be doing the wrong thing faster and more consistently.

We celebrate the “visible hands” of automation while overlooking the “invisible process knowledge.”

Manufacturing intelligence has never been built on impressive equipment alone.

It depends on getting every process and every cutting decision right.

02 | Process Know-How Is One of the Hardest Things to Make “Symbiotic” on the Shopfloor

In many factories, process knowledge still looks something like this:

• The tool is selected based on an experienced machinist’s intuition.

• Spindle speed and feed rate come from a process sheet written ten years ago — and no one dares to change them.

• The last tool failure was blamed on “hard material,” but no one can say which parameter actually crossed the limit.

• Whether a previous process plan can be reused next time depends on whether someone remembers it.

Strip away the anecdotes, and there are four very practical questions:

👉 Which tool should be selected for this part?

👉 Who decides the spindle speed and feed rate?

👉 Which parameter caused the last abnormality?

👉 Can the previous process plan be reused next time?

Getting a process decision wrong can cost far more than it first appears.

The cost of a wrong process decision =

Scrap × Part Value
+ Downtime for Tool Change × Hourly Production Value
+ Rework Hours × Labor Cost

The direct purchase cost of cutting tools typically accounts for only 3%–5% of total production cost in industry estimates.¹

Yet the impact of poor tool management — including inappropriate tool selection, inefficient tool usage, and stagnant inventory — can reach 20%–30% of overall cost impact through downtime, scrap, rework, and emergency procurement.²

The real cost is therefore not simply the tool itself.

It is what happens when the wrong tool is selected, the wrong parameters are used, or the right process knowledge cannot be reused.

The problem is not that experienced machinists are no longer valuable.
The problem is that experience cannot be scaled, cannot be present 24/7, and cannot make the right decision when the expert is on vacation.

03 | Let the AI Model Handle the Computation. Let Engineers Handle the Judgment.

So how can process knowledge move from “what an experienced machinist knows” to a capability that can actually be accessed and reused?

The first step is not to replace engineers.

It is to let AI take over the repetitive, time-consuming, highly standardized computational work.

This is where the Knowhy Process Agent comes in.

Upload a drawing, and the work can begin.

Given a part drawing, the Process Agent can help with some of the most repetitive tasks:

Reading drawings.
Checking tolerances.
Matching tools and cutting parameters.
Checking tool inventory and availability.

None of these tasks is particularly complicated.

But together, they consume a significant amount of an engineer’s time.

Now, these repetitive computations can be handled by an AI-powered process agent.

Engineers can move from “computation” to “judgment” — focusing on the decisions that truly require engineering expertise:

• Which process route is optimal?

• Where is there still room for cost reduction?

• Which abnormalities require expert judgment?

• Which process decisions can be standardized and reused?

Let the tools handle repetitive work.
Let engineers focus on judgment and creativity.

04 | Put the Process Agent to Work Before We Talk About Human–Machine Symbiosis

What Knowhy is building, at its core, is a way to turn every tool and every cutting operation on the shopfloor into usable, reusable data.

The Knowhy Process Agent uses that foundation to work from part drawings, handle high-volume process calculations, and support better cutting decisions.

Around it, the rest of the Knowhy cutting management ecosystem provides the necessary data foundation:

• Smart Tool Box — controls tool storage and circulation.

• Tool Lifespan Management — monitors tool life and supports precise tool-change decisions.

• SCB Smart CNC Box — monitors machine conditions and captures equipment data.

• Operations System — connects tools, machines, processes, and operational data into a unified foundation.

• Process Agent — turns accumulated data into actionable process intelligence.

The progression is simple:

Manage the tools →
Capture the data →
Build the knowledge base →
Put the data to work.

That final step is what turns data into decision-making capability.

The Real Question Behind Human–Machine Symbiosis

Even if you are not ready to deploy any system today, there is one question worth asking:

Does your factory’s process knowledge exist in people’s heads — or does it exist in the system?

Robots may give factories stronger bodies.

But intelligent manufacturing also needs a brain that can understand the process, learn from data, and support decisions.

Human–machine symbiosis is not just about machines working alongside people.

At its core, it is about sharing intelligence and decision-making.

And in machining, that starts with making process knowledge visible, reusable, and actionable.

Notes & References

¹ Estimates of cutting tool costs as a percentage of manufacturing cost vary by industry and production scenario. Research from Sandvik Coromant has cited tool costs at approximately 2%–4% of customers’ component manufacturing costs, while industry sources commonly reference a range of around 3%–5%.

² The 20%–30% figure refers to the potential overall cost impact associated with poor tool selection, tool usage, and inventory management, including indirect costs such as downtime, scrap, rework, and emergency procurement. The exact impact varies significantly by manufacturing process and operating conditions. The cost-of-quality references cited in the original analysis include benchmarks from ASQ and APQC.