How We 3D Print Custom Parts for Leo Rover: Materials, Process, and Tradeoffs

TL;DR: Fictionlab’s team uses 3D printed robot parts extensively across the Leo Rover platform – from sensor mounts and antenna brackets to enclosure lids and custom payload adapters. The default workflow relies on three filaments: PLA for prototyping and indoor fixtures, PETG for general outdoor use, and ASA for UV-exposed or thermally stressed parts. Each material trades printability against field durability, and choosing the wrong one usually shows up after the first week of dust, sun, or vibration. This article walks through how Fictionlab selects materials, designs for FFF printing, and what has actually broken in the field.

Why 3D Printed Robot Parts Make Sense for Leo Rover

The Leo Rover is an open-source mobile UGV, and a large share of its non-structural geometry is produced with fused filament fabrication (FFF). 3D printed robot parts let Fictionlab iterate fast, ship low-volume custom variants, and let customers modify the platform without retooling. When a research group needs a LiDAR riser, a GNSS mast, or an enclosure for a custom PCB, printing is almost always faster and cheaper than machining – as long as the part is designed with the limits of the process in mind.

That said, FFF is not free of tradeoffs. Layer adhesion is anisotropic, UV degrades many consumer filaments, and creep under sustained load is real. The choice of material and print orientation matters as much as the CAD itself.

What Materials Fictionlab Uses and When

Across years of Leo Rover deployments – including hot outdoor testing, agricultural pilots, and indoor lab use – three filaments cover most cases. Here is how they compare in practice, based on Fictionlab’s own field experience:

PLA vs PETG vs ASA: A Practical Comparison

The table below summarizes how each material behaves on the bench and in the field. Values are approximate and reflect typical FFF prints at 0.2 mm layer height with 40-50% infill.

• PLA – Easy to print, stiff, dimensionally accurate. Glass transition around 55-65°C, which means a closed rover chassis sitting in summer sun can deform PLA brackets. UV resistance is limited. Best for indoor prototypes, jigs, and parts that will be reprinted often.
• PETG – The workhorse. Tougher than PLA, better layer adhesion, glass transition near 80°C, acceptable UV resistance for several months of outdoor use. Slightly stringy and less crisp on fine details. Fictionlab uses PETG for many production accessory mounts shipped with Leo Rover.
• ASA – The outdoor choice. Excellent UV stability, glass transition around 100-105°C, and good impact resistance. Requires an enclosed printer, should be printed with good ventilation because of styrene-containing fumes, and warps without proper bed adhesion. Used for antenna mounts, GNSS masts, and any bracket that will live in direct sun for months.

A simplified decision matrix Fictionlab’s team applies:

1. Will the part live indoors and see no sustained load? Print PLA.
2. Will it go outside but stay in shade or be replaced seasonally? PETG.
3. Will it sit in direct sunlight, on a hot chassis, or near a motor driver heatsink? ASA.
4. Does it need chemical resistance, threads under cyclic load, or sub-millimeter tolerance? Consider machined aluminum, SLS nylon, or another suitable process instead.

How the Design Process Works for a New Leo Rover Part

A typical custom bracket for Leo Rover goes through four stages before it ships. The process is informal but consistent across the team.

1. Constraint Capture

Before opening CAD, Fictionlab’s engineers list the mechanical interfaces, the available mounting points on the top plate and body, the payload mass, expected vibration profile, and environmental class. For ROS or ROS 2 integrations this also includes cable routing for USB, Ethernet, or CAN – a bracket that blocks a connector is a bracket that gets reprinted.

2. CAD with FFF in Mind

Designing for FFF is different from designing for machining. Walls are sized in multiples of the nozzle diameter (typically 0.4 mm), overhangs stay under 45 degrees where possible, and load-bearing features are oriented so that stress runs in the layer plane, not primarily across layer lines. Heat-set brass inserts are preferred over printed threads for any joint that will be opened more than a few times.

3. Print and Inspect

First articles are usually printed in PLA, even when the final material will be different, because PLA is the cheapest way to confirm fit on the actual rover. If the geometry checks out, a second print goes in the target material. Layer height is usually 0.2 mm, with 0.12 mm reserved for parts with visible curves or shallow text.

4. Field Trial

Nothing replaces installing the part on a Leo Rover and driving it. Fictionlab’s team logs failures by mode: cracked layer line, stripped insert, UV yellowing, creep under load. Those failure modes feed back into the material choice for the next revision.

What Has Actually Broken in the Field

Behind-the-scenes, the lessons that matter most came from things that did not work. A few representative examples from Fictionlab’s field notes:

• PLA camera mount, summer field test: The mount sagged after several hours in direct sun on a dark chassis. Surface temperature measured with an IR thermometer was approximately 65°C. Switched to PETG, problem solved.
• PETG antenna bracket, multi-month deployment: Color faded and the part became brittle after roughly four months of continuous UV exposure. ASA was selected for the replacement.
• Printed M3 threads on a frequently removed lid: Stripped within a dozen cycles. Replaced with heat-set inserts, no further failures.
• Thin-wall enclosure for a power board: Cracked along a layer line after a drop test. Reoriented the print so the impact direction aligned with the layer plane, and added a 2 mm fillet at the stress concentration.

The pattern is consistent: most FFF failures are either material mismatch or print orientation, not the technology itself. 3D printed robot parts hold up well when the design respects the process.

When 3D Printing Is Not the Right Answer

FFF is not universal. Fictionlab moves to other processes when the requirements clearly exceed what filament can deliver:

• Machined aluminum for structural plates, motor mounts, and anything carrying the full rover mass.
• SLS nylon (PA12) for complex geometries that need more isotropic strength than FFF and finer features than FFF can hold.
• Sheet metal bending for large flat brackets where stiffness-per-gram matters.
• Injection molding for high-volume parts where tooling cost is justified by unit cost – generally not for one-off Leo Rover integrations, but relevant for repeat production components.

For one-off integrations and small batches, though, FFF remains the fastest path from sketch to working rover. That is why custom payload adapters for research customers almost always start as printed parts, and only graduate to other processes if volume or environment demands it.

How This Connects to Custom Robotics Projects

When research teams or integrators come to Fictionlab with a custom payload – a manipulator, a hyperspectral camera, a sampling arm – the first iteration of the mechanical interface is almost always printed. The same material logic applies: PLA to confirm fit, PETG or ASA for the deployed version, machined parts only where the printed version cannot meet the spec. This workflow is how custom robotics projects at Fictionlab move from concept to field-ready hardware in weeks rather than months.

FAQ

Can 3D printed parts survive outdoor robotics deployments?

Yes, with the right material. ASA and PETG hold up to months of outdoor use in Fictionlab’s testing. PLA is not recommended for exposed outdoor parts – it can soften in sun and degrades over time under UV.

Why not use carbon fiber filaments for everything?

Carbon-filled filaments are stiffer but more brittle, wear nozzles quickly, and rarely justify their cost for accessory-level parts. They make sense for specific stiffness-critical brackets, not as a default.

How does Fictionlab handle threaded fasteners in printed parts?

Heat-set brass inserts are standard for any joint opened more than a few times. Printed threads are reserved for one-time assemblies or very coarse pitches.

What layer height does Fictionlab use for Leo Rover parts?

0.2 mm is the default. 0.12 mm is used for visible surfaces or fine features. 0.3 mm appears on large structural prototypes where speed matters more than finish.

Are the Leo Rover printed parts open source?

Yes. The Leo Rover is an open-source platform, and files for standard printed components are published in the project repositories, with STL files and STEP/source files where available.

Can customers print their own replacement parts?

That is the intent. Customers regularly print replacement mounts, custom enclosures, and payload adapters on their own hardware. Fictionlab also prints and ships parts on request through the shop.

What printer hardware does Fictionlab use?

A mix of standard FFF machines with enclosed chambers for ASA. The workflow is intentionally compatible with hardware most labs already own, so customers can reproduce parts without specialized equipment.

Need a Custom Printed Part for Your Rover?

If a research or integration project needs a custom mechanical interface, sensor mount, or payload adapter for Leo Rover or Raph Rover, Fictionlab’s engineering team handles design, material selection, and production. Start a conversation through the custom robotics page.