Skip to content 
A manufacturer of heavy, safety-critical components assembled them across separate manual stations: five operators, hand tightening at torques up to 300 Nm, a scale, and a marking step before pack-out. GAV Sistemi consolidated the whole sequence into one automatic machine built around a rotary table, run by a single operator.
Assembly automation cases usually earn their keep on cycle time and headcount. This one is about mass. The part is heavy, the biggest joint runs at 300 Nm, and a new contract asked for volumes the manual line had no way to reach. When part weight, torque and volume climb together, the question changes shape: the whole line has to fit into one machine, and the machine has to carry everything the operators used to.
What did the manual line look like?
The original process was a chain of benches. An incoming part was checked and weighed. Two components were tightened together at a first station. The assembly then traveled by hand to the next bench, where a third component went on at 300 Nm. After that came a scale, a print whose content depended on the measured weight, and packing.
Each assembly weighs around 47 kg. Read that number against the layout and the line explains itself. Every move between benches was a lift, so moving the part was as much of the job as assembling it, and the distance between stations was set by what a person can carry, not by what the process needs. Five operators kept it running.
The 300 Nm joint was the hardest step to do consistently. At that torque the reaction has to go somewhere, and on a manual bench that somewhere is the operator’s arms and whatever the part happens to be braced against. The first joints of a shift and the last ones do not get pulled the same way. Tightening at this level is exactly the work a torque-controlled screwdriving station exists for; we walked through how those stations run, feeder to torque record, in our article on automated tightening. That article kept its numbers general on purpose. This case supplies them.
How does one machine absorb the whole line?
The machine is built around a rotary indexing table. Both tightening operations run on the table: the first joint at one station, the 300 Nm joint at the next index, pulled by a spindle instead of a person leaning on a bar. From there a handling robot takes the completed assembly through a dimensional check on a gauge, onto a scale, and past a printer that marks each part with its weight class. Then it sets the part on a cart, position one through six. When a cart is full, the machine calls the operator.
That call is the operator’s whole job on this machine: bring an empty cart, take the full one away. The six cart positions work as a buffer, so the machine keeps cycling while the exchange happens and one person covers the whole process without standing inside its cycle. The part travels on the table and in the robot’s gripper, never in anyone’s arms. At 47 kg per assembly, that path is a structural calculation in its own right: the table, the fixtures and the gripper are sized for the part from the first layout drawing.
| Step | Manual line | Automatic machine |
|---|---|---|
| First joint (parts 1+2) | Manual station | Rotary table station |
| High-torque joint (300 Nm) | Manual station, part repositioned by hand | Next index of the same table |
| Transfer between steps | Operators carry the assembly | Table indexing + handling robot |
| Dimensional check | Not systematic | Gauge station, every part |
| Weighing and marking | Separate scale, manual print step | In-cycle scale, class printed automatically |
| Pack-out | Manual packing | Robot loads carts; machine calls the operator |
| Operators | Five | One |
The rows follow the same order the benches did. The machine absorbed the process step by step, and where the manual line had relied on habit, it put a measurement.
Why is every part weighed and marked?
Because the downstream process depends on the exact weight of each unit. The weight bands are narrow, a few hundred grams apart, and a part in one band is handled differently from a part in the next. So the machine weighs every assembly in cycle and prints its class directly on the part. The information travels with the piece instead of on a sheet beside it.
On the manual line this was a separate bench with its own step. Separate steps drift: a scale reads, someone notes, someone else prints, and the piece and its number can part ways somewhere between benches. In cycle there is no gap for that to happen in. Every part gets weighed, every class gets printed, every result gets recorded. That is the same logic that drives marking automation in general: a mark earns its place when it is applied consistently and reflects a measurement you trust.
What happened to output?
Before the machine, the plant produced 8,000 parts in six months. A new supply contract then required 32,000 parts in a year, and the honest internal answer was that the manual line could not get there. With the machine running, the volume was delivered in under five months. Four times the previous half-year output, in less time than that half year took.
The numbers moved because the constraint moved. On the manual line, capacity was five people’s working hours, and a real share of those hours went into carrying, staging and repositioning a heavy part between benches. On the machine, capacity is cycle time: the table works both joints while the robot runs the checking, weighing and marking downstream, all in parallel. The economics of this kind of shift are covered more generally in Manual to Automatic Lines: ROI; this case is what the shift looks like on a part too heavy to treat casually.
What happens when the product changes diameter?
The contract alternates between two product diameters, and the machine converts between them. The fixtures and tooling for one diameter come off, the set for the other goes on, and the base machine, with its motion and its controls, stays. It has been converted back and forth as orders alternated. That behavior is designed in from the first layout: a changeover is engineered into the machine at the start, and it is one of the reasons a custom-built machine outlives the product it was first built for.
This is a different story from our rotary table case study, where the point was consolidating four manual stations and adding process control on every joint. Here the point is weight: of the part, of the torque, of the volumes. When all three are high, the gap between a manual line and a machine stops being a percentage and becomes the difference between refusing a contract and delivering it early.
Frequently asked questions
Can an automated station really tighten at 300 Nm?
Yes. Torque at that level is routine for a machine spindle and harder to deliver consistently by hand than almost any other assembly operation. The station applies the same torque on every cycle and records the result for every joint, on the first part of the shift and on the last.
Does the machine handle the part’s full weight?
Yes. The table, the fixtures and the handling robot are all sized for the part, so the assembly moves mechanically from load to cart. Operators handle carts; the machine handles the weight.
How is the weight-class mark used?
Each part leaves the machine carrying its own class, printed on the part itself. Downstream, the class determines how that unit is used, with no re-weighing and no paperwork traveling separately from the piece.