Gantry vs Six-Axis Robot for Machine Tending

Motionwell Automation builds both sides of the gantry vs robot machine tending question in Singapore, and the answer is settled by geometry before price. Our own linear gantry series runs a steel single beam with 10 to 30 m of X-axis travel at 50 to 200 kg payload, dual Z-axis heads servicing more than ten CNC machines in parallel. On the other side we integrate industrial arms across FANUC, ABB, KUKA, Yaskawa, Yamaha and JAKA platforms, and we design the low-profile seventh axis they ride on: 100 mm of overall track height at plus or minus 0.05 mm repeat positioning, delivered at 2,850 mm of stroke and 1.5 tonnes of platform load. Read as cartesian vs articulated robot, the choice is a question about the shape of your layout. Machines in a row favour a beam. Machines around a point favour a wrist. Machines are designed, assembled and tested at our Woodlands Link facility, and the company has delivered more than 150 special purpose machines since 2014 under ISO 9001:2015 and bizSAFE Level 3.

Where we stand, said plainly. The gantry is our own product series and the arm is something we buy, so we have a commercial reason to prefer the beam. That is why this page carries a full section on when a gantry is the wrong answer, and why that section names cases we would lose. We do not sell machine tools, we do not manufacture robot arms, and where a standard cell from your machine tool builder covers the job at a lower price we say so at concept review.

This page takes the decision in the order it gets made: geometry, orientation, floor against overhead space, machines served, cost per served machine, redeployment, maintenance, cycle-time structure, when a gantry is wrong, and the hybrid between the two. The wider build scope behind either choice, including machine interface signals, part presentation and swarf, is on our machine tending automation page; the arm-side work is on the robot integration services page; and the third family is our comparison of collaborative and industrial arms. If you have a machine list and a cycle time, skip ahead and talk to an engineer.

Are Your Machines in a Row or Around a Point?

This is the first question and it usually settles the shortlist, because the geometry decides everything downstream.

An articulated arm works inside a shell around a fixed base. Its reach is a radius, bought in every direction at once including the ones you will never use, and payload falls off sharply at full extension, exactly where a long layout asks the arm to work. More distance means a change of arm class, not another module.

A gantry works inside a rectangular prism drawn along a line. The X axis runs on rack and pinion in sections that can be joined without a fixed limit, so two more machines on the row means more beam and rail rather than another robot, another controller, another tooling set and another guarded footprint. Distance is the one thing a beam sells cheaply.

The test takes ten minutes with a layout drawing. Mark the load point of every machine, meaning the open door and the chuck or vice behind it rather than the machine footprint. Points on a line are covered by a beam, and every metre added is a module. Points on a circle around a spot where an arm could stand are covered from one base. Points scattered on both sides of an aisle suit neither, and the honest conclusion is one arm per cell, or a track down one side and a rethink of the layout.

What Does a Three-Axis Gantry Give Up at the Chuck?

Orientation. Every series in our linear gantry range moves in X, Y and Z, so the part enters the machine in the attitude the gripper picked it up in. The gantry presents the part as it holds it, and cannot tilt, roll or turn it on the way in.

Where blanks come from a magazine or a fixtured tray in the attitude they have to be loaded in and drop vertically into a chuck, that costs nothing. That is the case the beam was designed for. Where it stops being free is easy to check on a drawing: an angled or tilted fixture, a part that flips between the first and second operation, or a blank that arrives lying down and has to go in standing up.

A six-axis wrist solves all three directly, which is what the extra three axes are for. Three cheaper answers are worth pricing first: change how the blank is presented upstream, put a passive flip station between the pick and the place, or bring one arm in at the single station that needs the rotation and let the beam do the transport. Which is cheapest depends on how many part numbers need the flip, not on the robot.

There is a second-order effect worth naming. Because the gantry cannot fix orientation later, it has to be right at the pick, which pushes cost upstream into trays, magazines and locating fixtures, while an arm can accept a sloppier pick and correct on the way in.

Which Space Do You Actually Have, Floor or Overhead?

The two consume different space, and many plants we see are short of one and not the other.

A gantry travels overhead, so the floor under the row stays clear for the operator, the pallet truck and the swarf bin, and on a tight shop that aisle is often worth more than the difference in robot price. What it needs instead is clear height along the whole row and a structure to carry it: an overhead crane, service runs, lighting, extraction ducting and sprinklers occupy the same volume, and double-beam builds add a maintenance walkway and railings on top. Anything already up there is a conflict, cheaper to find on a photograph than on site.

An arm consumes floor: its own cell area plus safe access, and on a seventh axis a corridor down the row plus its guarding. The vertical cost of a track matters too, because a conventional seventh axis eats 200 to 400 mm of vertical working height before the arm has done anything. That is why we designed ours at 100 mm overall height with ground rack and pinion, plus or minus 0.05 mm repeat positioning and no floor pit, carrying ABB IRB-series arms as a coordinated external axis.

How Many Machines Can One Unit Hold, and What Stops It Adding One More?

One steel single-beam unit with dual Z-axis heads serves more than ten CNC machines in parallel on our own product record, with delivered lines running eight to ten machines from a single gantry. An arm covers one or two machines within reach, and more only on a track, where it visits machines serially the same way a beam does while still paying for a floor corridor and guarding.

What caps the number is arithmetic rather than mechanics. One unit holds as many machines as it can complete a full tending round trip for inside the shortest machine cycle on the row: divide that shortest cycle by tending time per machine plus travel, set out in full on our machine tending automation page. Two consequences follow. A row of fast machines is harder to serve than a row of slow ones, because tending time barely changes while the window shrinks, and a mixed row bottlenecks on whichever tool finishes fastest.

This is where the dual Z-axis head earns its cost. Two independent Z columns, each carrying a gripper pair, let one column lift finished parts out while the other puts blanks in, working four parts in a single visit rather than two trips. Our own product record for the series states the design cuts cycle time by 50 percent. Read that as what it is, a comparison of one visit against two on the tending sequence rather than a claim about your machine’s cutting time: its value to you is visits saved across a row.

Why Compare Cost Per Served Machine Instead of Cost Per Robot?

Because the two do not scale the same way, so one gantry against one arm compares nothing. The honest denominator is machines served.

Our own linear gantry product record states the position directly: one gantry comes in more than 50 percent below deploying multiple six-axis robots for the same job. That is a product-series position comparing one motion system against several rather than a quotation, and it holds for the work the beam is good at, a straight row of machines loaded vertically. Move any part of that premise and the arithmetic moves with it. The reason the gap exists is visible in what each option multiplies.

That last row matters more than it looks. The signals a machine tool has to give a robot, and the work of getting them out of a legacy control, are a cost per machine neither option avoids, so read the comparison as being about the motion system and the perimeter rather than the whole cell.

The trade also inverts sharply, because a cost per served machine collapses when the denominator is one. A beam, columns, full-length rail and cable carrier over a single lathe is the worst version of it, and an arm beside that machine usually wins on cost, footprint and installation time.

What Happens When the Shop Layout Changes?

The arm wins this one clearly, and it belongs in the comparison before the capital numbers.

A gantry is a structure sized to one row. Beam length, column positions, rail, cable carrier and drive sizing all follow the machines that were on the drawing when it was designed. Extending that row is modular, because the X axis is built in joinable sections. Moving the row is a re-installation: dismantling, re-levelling, re-aligning, re-commissioning, and whatever the guarding becomes in its new position.

An arm relocates with its pedestal and is re-taught, and a collaborative arm moves to another line and is re-taught in a shift. Our low-profile track relocates with it, because it sits on adjustable feet at 100 mm and needs no sunken pit.

Ask two questions before signing off a beam. How many times has this row been rearranged in the last five years, and is a product change coming that alters the machine mix? A shop that rearranges annually hands the capital saving back in re-installation. A row that has not moved since the building was fitted out is what the beam is for.

Which Is Easier to Maintain, and Who Has to Go Up There?

Maintenance access splits the two the opposite way from floor space, which is why it deserves more than a footnote.

A gantry keeps its mechanism above the process. Linear guides, ball screws and rack teeth collect chips and coolant mist strips grease, so a structure above the mess rather than in it is a genuine service advantage, and where the environment is dirty we use V-groove roller guides rather than recirculating ball guides, which is what our dual-rail truss manipulators use in dusty and harsh service, because a roller on a V edge sheds debris a ball guide would ingest. What it costs is a service point at height. Double-beam builds carry a maintenance walkway and railings for that reason; on single-beam builds, settle at design stage whether the Z head can be driven to a service position at one end of the row.

The cable carrier is the wear item the two do not share. A carrier running the full length of a 10 to 30 m beam and filling with hot chips becomes an abrasion problem long before it becomes a jam, so it needs continuous-flex cable rated for the travel, routed away from the chip conveyor discharge. An arm has a shorter dress pack and a smaller version of the same problem, covered from the tooling side on our end of arm tooling page.

An arm keeps everything at working height and is serviced without a ladder. What it costs is standing in the floor space beside the machine door, the space the machine tool’s own technician wants when the spindle needs work, and on a row that conflict repeats at every station. Design for fault recovery either way: after a stop, somebody has to get a part out of a machine with a gripper inside it, and a Z head over an open door in an unknown state is not a trivial extraction.

Why Is the Robot Idle Most of a Long Cycle, and Why Is That Correct?

Machine tending is the application where an idle robot is the right outcome, and misreading it is a common reason a sound project gets rejected.

The machine cycle sets the pace, and the robot’s only job is to finish its round trip inside that window, so on a long machining cycle it waits by design. That waiting is not waste. It is the headroom that lets one unit cover a row instead of one machine, and it is what you are buying when you buy a beam. An arm dedicated to a single machine converts the same idle time into nothing, which is why either layout is judged on machine utilisation and unattended spindle hours rather than on how busy the robot looks.

The comparison to run is therefore not top speed but tending time per machine plus travel, timed against the shortest cycle on your row. Where tending time itself is the constraint, the row has outgrown a single unit of either kind, and the answer is two units, a resequenced row or a buffer rather than a faster robot.

When Is a Gantry the Wrong Answer?

We build the gantries, so read this as the argument against our own product.

The machines are not in a line. A scattered layout, machines on both sides of an aisle, or cells placed by process rather than by geometry each push the answer back towards arms, because a beam has to span everything it serves. A beam spanning a room to reach four machines pays for structure that does no work.

One machine, or two. Cost per served machine collapses when the denominator is small. A beam, columns, full-length rail, cable carrier and a perimeter around a single machine buy structure that serves one door, where an arm on a pedestal is the smaller footprint and the shorter installation.

The part has to change attitude. Angled chucks, tilted fixtures, a flip between operations, a part that must be set down differently from how it was picked. Three axes present the part as they hold it, and no amount of travel substitutes for a wrist.

The overhead volume is taken. An existing crane over the row is the common one, along with ducting, sprinklers and a roof structure that will not carry a beam. A hard stop, and one to check before a concept drawing exists.

The layout will move. Covered above, and worth repeating because it is the failure that shows up two years after acceptance rather than at commissioning.

The mix changes faster than the parts. Where changeover cost dominates cycle time, a reprogrammable loading device beats a faster one. On our cleanroom automated test equipment build a dedicated gantry would have been faster per cycle, and it was still the wrong answer: mixed connector types and modest volume per variant meant the loading device had to be reprogrammable rather than quick, so a collaborative arm was specified instead, holding plus or minus 0.05 mm at the fixture interface. The reasoning is set out in the cleanroom test equipment case study.

The robot has to follow a path, not make a pick. Grinding, deburring, dispensing along a contour and anything with a process force in it are six-axis work. Our own example is force-controlled robotic grinding of CFRP composite panels holding 100 N constant contact force, developed with A*STAR SIMTech. A three-axis frame has no way to hold a tool normal to a curved surface.

The parts are light, planar and very fast. Metres of straight travel are cheaper and stiffer on a cartesian frame than on a rotary arm, but a short, fast, flat pick and place is SCARA territory rather than either machine on this page, as our SCARA robot selection guide sets out.

Two exclusions, said directly. We have not delivered a random bin picking cell, so if your blanks genuinely arrive heaped that is its own project with a proving trial rather than a line item on this one. We do not take on production welding cells, we do not issue CE certificates and we are not a notified body.

Is the Hybrid the Honest Answer?

Often, and it belongs on the shortlist as a third option rather than a compromise. The pure forms are the ends of a range, and many of the rows we are asked to quote sit inside it.

The first row is the one often left out of a two-way comparison. Buying a bigger robot to cover distance usually costs more than putting a normal robot on a track, and it does nothing for the multi-station case. Our track was designed for that job: 100 mm overall height so it does not eat the arm’s usable envelope, ground rack and pinion at plus or minus 0.05 mm repeat positioning, and integration with ABB IRC5 and OmniCore as a coordinated external axis, so track motion is planned with the arm rather than sequenced after it.

Two conditions decide whether a hybrid earns its extra interface. The work must be genuinely separable, which is a process question rather than a layout one. And somebody must own the handshake between zones: part present, zone clear, station ready, fault. An interface is where two suppliers can each assume the other owned it.

What Changes in the Safety Scope Between a Beam and an Arm?

The standards are the same for both. Start from an ISO 12100 risk assessment; the robot and the integrated system then fall under ISO 10218-1 and ISO 10218-2, with each safety function carrying a required performance level under ISO 13849-1, calculated and validated rather than asserted.

What differs is the shape of the guarded space and the hazards inside it. A gantry guards a long perimeter: fence length follows the row, and every point where an operator loads a magazine, clears chips or reaches a machine control is an access point to interlock or scan. It also carries a hazard the arm does not: a load travelling overhead where people stand. Where the machine incorporates lifting equipment it must be certified by an approved authorised examiner before use in Singapore. An arm guards a compact envelope with one or two access points, which on a track becomes a corridor with cross-traffic to manage. Scanner distances are calculated rather than chosen, from the ISO 13855 approach speed constant of 2,000 mm/s up to 500 mm and 1,600 mm/s beyond, plus stopping performance measured on the built cell.

One point applies to both and is easily missed: the machine tool has its own door interlock and stop category and the cell has its own, so somebody must own that boundary and prove the robot cannot enter while the spindle can turn. Singapore does not require CE marking, but the Workplace Safety and Health Act requires dangerous parts of machinery to be securely fenced. We deliver the guarding scope that follows either choice, described on our machine safety and CE marking page.

Which One Should You Specify?

Run these in order. The first that gives a hard answer usually settles it, and where two disagree the answer is normally the hybrid above.

1. Where do the load points sit? Mark the open door and the chuck for every machine. A line points at a beam, a circle at an arm, a scatter at neither.
2. Does the part change attitude between the pick and the place? If it does, either buy the wrist or design the flip, and price both.
3. What is above the row? Clear height with the crane, ducting and sprinklers counted in. A hard stop, not a preference.
4. How many machines, and what is the shortest cycle among them? Machines served is that shortest cycle divided by tending time plus travel, and it is what makes cost per served machine mean anything.
5. How often does this layout change? Annual rearrangement argues against a fixed structure regardless of everything above it.
6. How many part families share the row? High mix with frequent changeover moves the argument towards a reprogrammable arm.
7. Who services it, and how do they reach it? Work at height on a beam, floor space beside the door on an arm.

Lead time runs 16 to 24 weeks from concept approval to factory acceptance testing on a standard build. Design, fabrication, assembly and testing happen at Woodlands Link with an in-house design team of eight, which is why a Singapore buyer attends the factory acceptance test instead of flying to it. That matters on tending work, because the details that decide whether a cell runs are gripper and fixture details found on your own parts. The delivered gantry work sits in context on our warehouse and intralogistics page.

Frequently Asked Questions

Is a gantry really cheaper than several six-axis robots?

On our own linear gantry product record, yes, and the framing matters: that record states one gantry costs more than 50 percent less than deploying multiple six-axis robots for the same job. It is a product-series position comparing one motion system against several, not a quotation for your row, and it assumes the job a gantry is good at, which is a straight line of machines loaded vertically. Cost per served machine is the number worth comparing. A beam over ten machines divides its cost ten ways; a beam over one machine divides it once, and there an arm usually wins.

Can a three-axis gantry load a part into an angled fixture?

Not as it stands. Every series in our linear gantry range moves in X, Y and Z, so the part enters the machine in the orientation the gripper picked it up in. If the chuck sits at an angle, if the part flips between the first and second operation, or if blanks arrive lying down and have to go in standing up, that orientation change has to live somewhere: in how the blanks are presented, in a passive flip fixture, or in an articulated arm at the one station that needs it. A six-axis wrist buys the freedom directly, and pays for it in reach.

What happens to a gantry if we rearrange the shop floor?

You re-install it, and this is the clearest advantage an articulated arm holds. A gantry is a steel structure sized to one specific row: beam length, column positions, rail, cable carrier and drive sizing all follow the machines that were there when it was designed. Adding two machines to the end of a row is a modular extension. Moving the row is not. An arm on a pedestal relocates and is re-taught, and our low-profile linear track is 100 mm high on adjustable feet with no floor pit, so it travels with the arm. If your layout changes yearly, weigh that before the capital comparison.