On this page
Flexible material handling is a range problem, not a part problem — the tool has to be chosen for the hardest part in the family, not for the part in the photograph. This single reframing prevents most of the failures in this application. Cells that handle a range of products badly are almost never defeated by the average part. They are defeated by the smallest one, the most fragile one, or the one with the worst surface, and that part was never in the selection conversation.
The second important idea is that flexibility is not the same as universality. A gripper that can be programmed to different positions and forces covers a wide range within limits set by its finger geometry and its force resolution. Beyond those limits no amount of programming helps, and the honest answer is a second finger set rather than one tool stretched past its useful range.
Start by defining the range properly
Most selection errors in flexible handling trace back to an incomplete description of what has to be handled. The description needs to be a range, and it needs to name the extremes.
| What to define | Why the extreme matters more than the average |
|---|---|
| Smallest and largest part dimensions | Sets the stroke range. The smallest part also limits how much contact area is available for friction |
| Lightest and heaviest part | Sets the force range and, with the acceleration, the friction requirement |
| Most and least rigid part | Sets the upper force limit. A range defined by the rigid parts will crush the compliant ones |
| Best and worst surface condition | Determines pad material. An oily or dusty part can defeat a pad that works elsewhere |
| Most and least consistent presentation | Determines whether the gripper can succeed alone or needs help from a feeder |
| Hottest and coldest part | Rules finger materials in or out. Most pad polymers have a narrow comfortable band |
Writing this list out takes twenty minutes and typically changes the answer. The recurring discovery is that the smallest part is harder than the largest, because it offers the least contact area and the least margin against slipping.
Define the force window before choosing the tool
Flexible handling needs two force numbers, not one. The lower number is the minimum that holds the part securely through the worst acceleration in the cycle. The upper number is the maximum the part tolerates without permanent change. A tool that cannot be set anywhere between those two values is not suitable, however capable it looks on paper.
The window is often narrower than expected. A thin-walled housing may tolerate very little compressive load at its middle but considerably more at a ribbed edge, and that difference alone can decide whether the right answer is a softer pad, a different contact location, or a different handling method entirely. Where the window collapses to almost nothing — the part deforms before it is held securely — the application needs a supporting method rather than a gripping method. How the lower number is derived is set out in gripping force: what it means and how to specify it.
Contact geometry decides whether the part survives
Two grippers applying the same total force to the same part can produce very different results, because what damages a part is local pressure rather than total force. Spreading the same load over twice the area halves the pressure.
| Contact approach | Pressure distribution | Suits | Poor fit for |
|---|---|---|---|
| Hard flat pads | Concentrated at the highest points, sensitive to surface flatness | Rigid parts with machined faces | Compliant or finished surfaces |
| Soft elastomer pads | Conforms, spreads load, forgiving of uneven faces | Fragile and finished surfaces, moderate force | Hot parts and abrasive or oily surfaces |
| Contoured or machined-to-part fingers | Even, predictable, repeatable | A narrow family of parts with stable geometry, high volume | Wide part ranges, where the contour matches only some variants |
| Multiple small contact points | Locally high, but tolerant of irregular geometry | Irregular or rough parts needing a stable three-point style location | Soft surfaces that take a set under point loads |
| Support from below rather than side grip | Distributed across the base | Deformable items with no reliable side surface | Parts that must be released into a constrained nest |
The choice between a two-finger and a three-finger arrangement also sits here, because it changes how the part is located as well as how the load is spread. The trade-offs are compared in parallel gripper vs 3-jaw gripper.
What wide stroke costs you
Wide stroke is the obvious answer to a wide part range, and it is usually part of the answer. It is worth being clear about what it costs, because those costs are what make an over-specified tool perform worse than a smaller one in the same cell.
- Mass at the wrist. A longer stroke generally means a longer body or a heavier mechanism, and every gram at the tool reduces the payload available for the part.
- Finger length and deflection. To use a wide stroke on a small part, the fingers have to reach further, and longer fingers deflect more under load, which shows up as placement error.
- Force resolution at the low end. A tool sized for large parts may not be able to resolve the small forces a fragile part needs with enough precision to be repeatable.
- Cycle time. Moving a long stroke from open to closed takes time, and in a mixed cell where every part needs a different opening, that time is paid on every cycle.
The practical compromise on many cells is a moderate stroke with interchangeable finger sets, rather than the widest stroke available. Where the part family genuinely spans a large width range and changeovers are expensive, a long-stroke tool earns its cost — but the decision should be made knowingly. For a mixed range that stays within moderate sizes, a compact parallel gripper with programmable force is often the more precise starting point.
Recipe strategy for mixed product families
Programmability only pays if it is used deliberately. A cell that can store recipes but runs everything on the default force and speed is not a flexible cell, it is a cell with unused capability, and it will fail on the parts that needed different settings.
| Element of the recipe | What it should be set from | Common omission |
|---|---|---|
| Open position | The part width plus a clearance that clears the nest | Reusing the previous recipe’s opening and accepting a slow approach |
| Closed position or force limit | The part’s deformable limit, not the gripper’s capability | Setting one force for the family and crushing the compliant variants |
| Approach and close speed | How much the part moves if it is disturbed on contact | Running everything at one speed for simplicity |
| Holder criteria | What a correct grip looks like in position and force terms | Relying on a commanded position with no confirmation |
| Finger set in use | The part family, verified at changeover | No check that the physical finger set matches the loaded recipe |
The last row is where costly mistakes happen, because a mismatched finger set can produce a grip that looks correct and damages the part slowly. A verification step at changeover is worth more than any amount of force resolution. For cells built around cobots, the constraints of small and varied parts are covered further in best cobot gripper for small parts handling.
Surfaces that change the answer
Surface condition is the most under-specified variable in flexible handling, and it can invalidate an otherwise sound selection. A pad that holds a clean steel part securely may slip on the same part with a light oil film, and a pad that works on a dry part may fail entirely once a fine dust layer is present.
- Oily or coolant-covered parts. Friction drops sharply and elastomer pads may swell or harden over time. Plan for a harder contact material and a higher force margin, or clean the part first.
- Dusty or powdery surfaces. Particles become a low-friction interface. Contoured fingers that push contamination aside work better than flat pads that trap it.
- Hot parts. Temperature limits narrow the pad material choice before it affects anything else.
- Finished or cosmetic surfaces. Here the constraint is not holding the part but not marking it, which usually means the contact material must be softer than the part surface.
When orientation has to come before gripping
A significant share of flexible handling cases turn out to be presentation problems wearing a gripping costume. If the parts arrive in a random orientation, no end effector can reliably pick the correct face, and the cell needs vision, a feeder or a mechanical orienting device before the gripper can contribute anything.
The diagnostic question is simple: at the pickup point, is the part’s gripping feature in a known position? If yes, the gripper is the right place to invest. If no, investing in the gripper is spending money on the wrong constraint, and the cell will continue to underperform.
The limits of programmable force control
Programmable force is genuinely useful and it is worth being precise about what it does not do. It does not measure the force actually applied to the part, it controls the force the drive generates. Between the drive and the part sit the mechanism, the finger geometry and the contact pad, and each of them changes the relationship between the commanded value and the load on the surface.
This means two things in practice. A force setting derived on one part geometry does not transfer directly to a different geometry without verification. And a force limit is not a safety guarantee for a fragile part — it is a repeatable starting point that has to be confirmed on the real product. Where the part is valuable or the consequence of damage is high, a first-off check after a change is cheaper than a scrap event.
Frequently asked questions
What makes material handling flexible?
The cell can absorb variation in size, shape, surface or product mix without major mechanical changes. In practice that comes from programmable position and force settings plus finger geometry that covers the range — both are needed.
Is a wide-stroke gripper always better?
No. Wide stroke covers a larger part range but usually adds wrist mass, needs longer fingers with more deflection, and can be less precise at low force. A moderate stroke with swappable fingers is often the better answer.
Can electric grippers handle fragile parts?
Yes, when the force can be limited below what the part tolerates and the contact pressure is spread over enough area. Both conditions have to hold; force control alone does not protect a part from a concentrated contact.
Do I need custom fingers?
Usually yes for a mixed part family. The gripper body sets the capability but the fingers decide where and how hard the part is contacted, and that is what determines whether the range works.
Why does my grip work on one part and slip on another of the same size?
Surface condition is the usual reason. Oil, dust, temperature and surface finish change friction far more than small dimensional differences, and a part that looks identical may present a very different contact.
What information makes a selection review useful?
The full part range with photos, weights and dimensions for the extremes, surface and temperature conditions, the presentation at the pickup point, the robot model and payload, the target cycle time, and any surfaces that must not be marked.
Image reference
