Direct answer: Mobile robotic welding is justified when moving the equipment is cheaper and safer than moving the part. Confirm that the member can stay where it lies, that the platform has a route and a stable stop, and that the seam can be divided into sections a single stop can reach. Floor condition, access, power, and sectioning belong in the plan before any equipment is selected.
Who this is for: This guide is written for fabrication managers and project engineers deciding whether a mobile welding platform belongs in an existing shop, and for the people who will have to operate it.
Scope: It covers how to plan mobile robotic welding from the part, the access route, and the sectioning plan. It does not cover welding procedure development, platform navigation technology, or the structural design of the assembly.
Mobile robotic welding starts with a question about the part
Fixed cells assume the part comes to the robot. That assumption quietly carries a crane, a fixture, a transport route, and a handling risk. For heavy fabricated members, long assemblies, and site work, the assumption is often the most expensive part of the concept, and it is worth testing before any equipment discussion.
In the reference footage for this guide, a long member rests on trestles on the shop floor while a mobile platform carrying the arm and the power source is positioned next to it. That is the whole idea in one frame: the part stays, the equipment arrives. Whether that is the right choice depends on how often the part would otherwise be lifted, and what that lift costs in time, risk, and floor space.
Access route, stop position, and stability
A mobile platform is only as good as the route it can take and the stop it can hold. Floor flatness, thresholds, cable crossings, drainage channels, and turning space decide where the equipment can actually stand. A stop that is reachable but not stable is worse than no stop at all, because it silently changes the relationship between the arm base and the joint.
Plan the stop positions on the layout drawing, not on the day. Mark where the platform stands for each seam section, where cables and gas run, where the operator stands, and where the crane path crosses. Sites that skip this step usually discover the conflict after the platform is already in the aisle. According to ISO 12100:2010, hazards are identified for the task as performed rather than for the machine in isolation, which is why each planned stop is treated as its own working situation.
Sectioning: one stop, one reachable set of joints
From any stop, the arm covers a bounded set of joints. That is why a mobile welding plan is really a sectioning plan: the seam population is divided into sections, each section is assigned to a stop, and the order of stops is chosen so the assembly stays dimensionally sensible while it is being welded.
Sectioning also drives the datum question. Every time the platform moves, the relationship between the robot base and the workpiece changes, so the working frame has to be re-established before the arc starts again. Treat that re-establishment as a named step with its own acceptance evidence, not as an assumption. The same discipline appears in datum recovery for large structures, where relocation is planned rather than improvised. EVST asks for the section order and the datum-recovery step in the same document, because one without the other is not a plan. According to ISO 3834-1:2021, the quality requirement level is chosen for the welding operation as a whole, which is why a sectioned seam is still governed by one selected level across every stop.
Decision table: fixed cell, mobile platform, or both
Notice that the mobile rows do not remove work; they relocate it. The engineering effort moves from fixture design to access, stops, and datum recovery.
| Situation | Usually points to | Main planning burden |
|---|---|---|
| Repeating part, moderate size, stable volume | Fixed cell with fixture | Fixture repeatability and loading ergonomics |
| Heavy or long member that is difficult to lift | Mobile platform | Access route, stop stability, and datum recovery at each stop |
| Mixed families with long and short members | Fixed cell plus mobile platform | Deciding which family goes where, and keeping one welding procedure set |
| Work performed where the assembly is built | Mobile platform | Site power, extraction, safeguarding of a moving work area |
| Seam population not yet mapped | Hold the decision | Complete the seam map and sectioning plan first |
Safeguarding a work area that moves
A moving work area is harder to guard than a fenced cell, because the boundary changes with the platform. ISO 12100:2010 provides the general framework for hazard identification, risk estimation, risk evaluation, and risk reduction, and it is the right starting point when the guarded space is defined per stop rather than once.
Integration, commissioning, operation, maintenance, and decommissioning requirements for robot applications and cells are addressed by ISO 10218-2:2025. Applied to mobile welding, the practical questions are how the safeguarded area is re-established at each stop, how access is controlled while the arc is live, and how the platform is brought to a safe state when work is interrupted.
EVST treats those answers as deliverables of the plan, alongside the welding procedure, because they determine how many stops per shift are realistically achievable in a shared shop. According to ISO 10218-2:2025, integration and commissioning requirements apply to the complete robot application, which is why a moving work area is specified with the same rigour as a fenced installation.
What acceptance looks like for mobile welding
Acceptance for a mobile application is not a single demonstration weld. It is a representative section welded from a representative stop, with the datum re-establishment step included, and with the quality requirements chosen in advance. ISO 3834-1:2021 sets out criteria for selecting the appropriate level of quality requirements for fusion welding, and naming that level early keeps the trial meaningful.
Record the interruption behaviour as well. If the platform loses position, or the work is stopped mid-section, the part identity, the completed sections, and the revalidation steps all have to be traceable. That record is what makes the second and third project easier than the first, and it is the difference between an experiment and a process.
What changes in the welding procedure documents
A mobile application does not change the physics of the joint, but it does change the conditions recorded around it. Position, access, and the state of the assembly at the moment of welding are all part of the procedure, and all three vary with the stop being used. That is why the procedure set for mobile work usually names the section as well as the joint.
Preheat and interpass conditions deserve explicit attention on large fabrications, because a member that stays on the shop floor is exposed to ambient conditions that a fixture in a fixed cell would partly control. If the procedure depends on those conditions, the plan has to say how they are established and checked at each stop rather than assuming a cell environment.
Distortion control also moves into the sectioning plan. The order in which sections are welded is not only a reach question; it decides how the assembly moves while it is being joined. A section order that is convenient for the platform and wrong for the structure is a common early mistake, and it is avoided by planning both together.
Where a mobile platform is usually the wrong answer
Mobile platforms are not a universal upgrade. Small repeating parts are almost always better served by a fixed installation with a good fixture, because the fixture repeatability that a fixed station provides is exactly the thing a mobile application has to re-establish at every stop.
Shops with congested aisles, uneven floors, or heavy shared crane traffic often find that the platform spends more time waiting for a clear route than welding. In that situation the honest conclusion is that the constraint is the shop, not the equipment, and that floor and layout work has to come first.
Applications that need a continuous long seam welded in a single uninterrupted pass are also a poor match, because sectioning is inherent to the method. Where an uninterrupted pass is a requirement of the procedure, a fixed installation with an external axis usually remains the better answer.
Utilities, extraction, and the shared shop
A mobile platform brings its own arm and usually its own power source, but it does not bring its own supply. Power capacity at the point of use, the route for cables that cross an aisle, gas supply and cylinder handling, and fume extraction that follows the work all have to be planned as part of the application rather than solved on the day.
Extraction deserves particular attention because it is the item most often left to improvisation. A fixed cell can have a hood designed once. A mobile application moves, so extraction either travels with the platform, exists at every planned stop, or is handled by general ventilation with a stated justification. Choosing between those three is a design decision with cost and layout consequences.
The shop itself is the other shared resource. If the platform stands in an aisle, it competes with forklifts and cranes; if it stands inside a bay, it competes with the assembly work happening there. Mapping the stops against the shop’s normal traffic pattern usually reveals a small number of positions that work well and a larger number that only work when the shop is quiet.
None of this argues against mobile welding. It argues for planning it as a shop project rather than an equipment purchase, which is also why the plan should name who owns the route, the utilities, and the stop positions after handover. EVST scopes this work as a shop project rather than an equipment purchase for exactly that reason.
Frequently asked questions
Is a mobile welding robot less accurate than a fixed cell?
The comparison is not accuracy against accuracy; it is where the reference comes from. A fixed cell inherits its reference from a fixture, while a mobile application re-establishes a working frame at each stop. Plan and prove that step and the application behaves predictably.
Do we still need fixtures with a mobile platform?
Often yes, but smaller ones. Local clamping, stops, and tack sequencing still control distortion. What usually disappears is the large transport-and-locate fixture used to bring the whole member into a cell.
How many stops should a section plan have?
As few as the reach and the assembly sequence allow. Every extra stop adds a datum re-establishment and a safeguarding change, so consolidating sections is usually more valuable than shortening cycle time within a section.
What limits mobile welding in an existing shop?
Usually the floor and the aisle, not the robot. Flatness, thresholds, crossing services, crane routes, and shared access decide whether the platform can stand where the seam is.
Project inputs for an application review
Send the following and the mobile-versus-fixed question can be answered against your shop rather than in principle:
- the structure drawing and the seam map with intended section boundaries
- member mass and how it is supported during welding
- floor and aisle conditions along the intended route, including thresholds and services
- available power, gas supply, and extraction at the planned stops
- the quality level and inspection method you have to meet
Send the structure drawing, the seam map with section boundaries, the member mass and support method, the floor and aisle conditions on the intended route, the available power and extraction, and the quality level you must meet. That set is enough to review whether a mobile welding platform fits the work. Related reading: automatic welding system solutions, datum recovery for large structures, large workpiece pose and reach testing, steel structure welding applications.