How a Painting Robot Is Integrated into a Paint Line in Indonesia

The reach study: proving the envelope before the robot is ordered

The first step in integration is always a simulation, never a purchase. Part geometry, booth dimensions, conveyor direction, and the robot mounting position go into the simulation, and the spray motion is run on screen.

What is being looked for is not only whether the applicator tip can touch every surface. How it gets there matters just as much. A position can be reachable and still force the arm through an extreme posture — close to an axis limit — and in that posture motion slows, the applicator angle to the surface shifts, and the film thins in that area even though the spray parameters never changed.

The second thing checked is clearance. A booth contains exhaust, filters, framework, hangers, and sometimes an old reciprocator nobody has removed. The simulation checks clearance along the whole path, not just at the end points.

The output of a reach study is a set of binding decisions: which MPX Series model, where it stands, and whether it needs a rail. Changing any of the three after installation costs far more than repeating the simulation.

• Reach to every surface that must be sprayed, undersides and recesses included
• Arm posture along the path, not only at the destination points
• Clearance to booth walls, exhaust, hangers, and the parts in front and behind
• A cycle time achievable without demanding maximum speed on every move
• Mounting choice — floor, rail, or wall — and what it does to the envelope

Offline programming and teaching: when each is used

There are two ways to write a painting robot program, and most projects use both in turn. Offline programming builds the path from a 3D model of the part on a computer, without stopping production. Teaching happens on the floor with the teach pendant, using the real part.

Offline wins for new parts whose model already exists and for work that must not halt the line. Its weakness is that a 3D model is never quite reality: hangers swing, parts sit slightly askew, and jig fabrication tolerances join in.

So the order we use is almost always the same: the path is built offline, then verified and refined on the floor with real parts. What gets corrected in the second stage is usually not the shape of the path but the applicator's distance and angle at a handful of points.

For plants whose parts change often, being able to write programs in-house is worth more than the number of programs we hand over at the start. That is why programming training sits inside the commissioning scope rather than being sold as a separate service.

The signals between robot, booth, and conveyor

A painting cell works when every part of it agrees on one sequence: the part enters, the part is in position, the robot may spray, the robot has finished, the part leaves. The signalling that enforces that agreement is the most underestimated piece of integration and the most common source of trouble.

On a continuously moving conveyor, the robot needs to know not just that a part is present but where it is now and how fast it is travelling. That capability is conveyor tracking: position data from the conveyor encoder lets the robot shift its path to follow the part, so a path taught on a stationary part stays correct on a moving one.

On an indexing conveyor the requirement is simpler: confirmation the part has come to rest, and permission to start. Yet this is where mistakes cluster, because permission is often given before the hanger has actually stopped swinging.

The booth carries signals of its own. Insufficient exhaust, an open door, or a drop in purge air pressure all have to be able to stop the robot before any paint leaves the applicator. That circuit is proven point by point before the cell ever runs with paint, not afterwards.

• Part presence detection and position confirmation before spray permission is granted
• Conveyor position and speed data for tracking a moving part
• Part type identification so the right program is selected automatically
• Interlocks on booth exhaust, access doors, and purge air pressure
• A ready signal from the paint supply: pressure and flow at the applicator
• Alarm feedback to the HMI so the operator can see what stopped the cell

Purge air, clean supply, and cable routing inside the hazardous area

A painting robot holds the inside of its body at a higher pressure than the booth air, using clean air, so solvent vapour cannot reach the electrical components inside. This system is what allows the robot to stand in the booth at all, and it is what deserves the most attention during installation.

Three things determine how good it is: the air source must be genuinely dry and clean, the pressure must be monitored continuously, and a pressure drop must stop the robot automatically. Plant air carrying moisture or oil does not just reduce reliability; it removes the compliance basis a certified robot arrived with.

Controller placement follows the same rule: the controller lives outside the hazardous area, and the cable between controller and robot runs along a defined route, sealed at the penetrations. Improvised cable routing is one of the audit findings we most often have to correct on cells installed without planning.

Paint and air lines to the applicator are routed at the same time but for different reasons: the hose has to follow the robot's motion without snagging, without sharp bends, and without surplus length, because every extra metre is more material wasted at every colour change.

Trajectory tuning: from correct motion to a correct finish

A robot with a correct path does not automatically lay an even film. Thickness at any point comes from the applicator's distance to the surface, its angle, the speed of travel, and how much the spray patterns overlap. Setting all four is its own job once the path is done.

What goes wrong most often is speed through corners. The robot slows as it turns, and if the spray keeps running at the same flow, paint builds up there. So paths around part corners are usually split, or material flow is made to follow the travel speed.

The next mistake is inconsistent pattern overlap. The spacing between passes has to stay constant across the surface, and on a contoured part that spacing drifts if the path only follows a flat projection. This is where most of the tuning time goes, and where a robot genuinely outperforms a manual operator.

Verification is done by measuring film thickness at agreed points, not by judging appearance. Those points are set with your quality team before tuning starts, so the result can be compared against the same specification production is judged by.

• Applicator distance to the surface held constant, contoured areas included
• Applicator angle square to the surface rather than skewed along the pass
• Travel speed and material flow kept in step, corners and turns above all
• Pass spacing that produces even overlap across the whole surface
• Film thickness measurement points agreed before tuning, not after

Colour change: the piece that sets the daily cost

In a robot cell, a colour change is not swapping a paint can. It means flushing the whole run between the change point and the applicator, and the length of that run decides how much paint is wasted and how long production stops every time the colour changes.

So a decision taken at the design stage lands every single day for the life of the cell: how close the colour changer sits to the applicator, how many colours are provided for, and how much material stays behind in the line after flushing.

Plants changing colour several times a shift usually benefit from a colour changer mounted on the robot arm, as close to the applicator as possible. Plants running one or two fixed colours have no reason to pay for that complexity.

The figure worth tracking once the cell is running is not just total paint consumption but material wasted per colour change. That number is usually what shows whether the flushing procedure still has room to improve.

Routine maintenance and the symptoms not to ignore

A properly maintained painting cell rarely stops without warning. Almost every major fault signals first, and it signals in the same few places.

Purge air pressure sagging slowly usually means a leak in a seal or in the air line, and leaving it means letting the cell run outside its compliance basis. A paint hose turning stiff or changing colour means material hardening inside it, which sooner or later releases as particles in the finish.

On the robot itself, what needs scheduling is grease replacement and checking the encoder battery — losing position data forces a recalibration and stops production for longer than the maintenance ever would. At the applicator, nozzle and air cap wear develops slowly and usually shows up first as a spray pattern that has widened.

We advise plants to hold their own stock of the consumables whose replacement is predictable — nozzles, seals, filter elements, and encoder batteries — because it is precisely these small components that most often stop production while a shipment is awaited.

• Purge air pressure monitoring and clean air quality checks
• Paint and air hose inspection for wear, stiffening, and leaks
• Scheduled grease replacement and encoder battery checks on the robot
• Nozzle and air cap treated as consumables rather than as failed parts
• Logging paint consumption and waste per colour change as an early indicator