RobotAIGeek

ABB Robotics Launches Flexley Stack F712, an Autonomous Forklift That Navigates by Sight

ABB Robotics launched the Flexley Stack F712, an autonomous forklift that uses camera-based vSLAM navigation instead of floor markers or beacons, bringing shared 3D fleet mapping to warehouse and automotive operations.

luna
4 min readPosted: Jul 8, 2026
ABB Robotics Launches Flexley Stack F712, an Autonomous Forklift That Navigates by Sight

Source: The Robot Report (Tier 1) · Story: ABB Robotics (Zurich, Switzerland) — satisfies the "Europe country" preference.

Content image (in-article): https://www.therobotreport.com/wp-content/uploads/2026/07/F712-Visual-SLAM-_-AMR-Studio-featured.jpg

Thumbnail image (distinct from content image): https://images.unsplash.com/photo-1581091226825-a6a2a5aee158

ABB Robotics has unveiled the Flexley Stack F712, an autonomous forklift that swaps warehouse markers and reflectors for a camera-based navigation system born out of the company's Swiss robotics research arm. Unveiled on July 7 and rolling out globally, the F712 is the latest piece of ABB's push into what it calls "autonomous versatile robotics" (AVR) — machines built to operate on messy, ever-changing warehouse floors instead of along fixed, pre-mapped lanes. For a company headquartered in Zurich, Switzerland, the launch also doubles as a statement: Europe's industrial automation giants still set the pace in physical AI, even as attention in the sector gravitates toward humanoid robots and Silicon Valley labs.


What Makes the F712 Different

Traditional automated forklifts typically depend on installed infrastructure — magnetic strips embedded in the floor, reflective tape on the walls, or dedicated beacons scattered around a facility. The F712 skips all of that. It uses visual simultaneous localization and mapping, or vSLAM, to build and continuously update its own three-dimensional understanding of a warehouse as it drives. ABB says the system achieves positional accuracy of ±10 mm, tight enough to slot pallets into narrow racking without retrofitting the building first. The forklift handles open and closed pallets, containers, and racks, lifting loads up to 2,000 kg (4,409 lb) to heights of 8.5 meters (27.8 feet), and it can travel at up to 1.7 meters per second (3.8 mph) while fully loaded — brisk for a machine sharing floor space with people and other equipment.


Vision Over Infrastructure: How vSLAM Works

The navigation stack traces back to Sevensense, an ETH Zurich spinout that ABB absorbed into its robotics division specifically to scale up vSLAM across its mobile robot lineup. Where older systems stitch together a flat map from thin slices of lidar data, ABB's approach builds a dense 3D picture of the environment and shares it across an entire fleet of machines, letting stackers, movers, and tugging robots all reference the same live map instead of operating in isolation. That shared spatial awareness is what allows different robot types on the same floor to coordinate rather than simply avoid each other.

"Traditional localization looks at the world with thin 2D lidar slices, but now we have a rich 3D view and can share maps across our fleet of stackers, movers, and tuggers. This is a turning point in industrial robotics." — Renaud Dubé, Head of R&D, Sevensense by ABB

Built to Work as a Fleet, Not a Standalone Machine

The F712 joins the existing Flexley Tug and Flexley Mover in ABB's vSLAM-equipped lineup, and all three are now VDA 5050-compatible, the industry standard that lets autonomous mobile robots from a shared ecosystem talk to one another regardless of task. Everything runs through ABB's AMR Studio, a no-code, drag-and-drop fleet management suite that the company says can cut commissioning time by up to 20% and adapt instantly when a warehouse layout changes. The F712 is certified to current ISO and ANSI safety standards and is aimed squarely at automotive and industrial customers dealing with line feeding, end-of-line handling, body- and press-shop logistics, and general warehouse storage and retrieval.

The bigger signal here isn't the forklift itself — it's the labor argument ABB is making around it. Marc Segura, president of ABB Robotics, framed the launch around hiring pressure in warehouses: businesses need to move more goods faster with fewer available drivers, and a machine that can be deployed without infrastructure changes lowers the barrier to automating that gap. Whether vSLAM-based AMRs actually spread faster than infrastructure-dependent systems will depend less on the sensor technology and more on how well fleets like this integrate into warehouses that already run a patchwork of equipment from different vendors.

Quick Q&A

Q: What is the ABB Flexley Stack F712?
A: It's a new autonomous mobile forklift from ABB Robotics that uses camera-based vSLAM navigation instead of floor markers or beacons, designed for warehouse and automotive material handling.

Q: How is vSLAM different from how older automated forklifts navigate?
A: vSLAM uses cameras to build a live 3D map of the environment in real time, rather than relying on pre-installed lidar reflectors or magnetic strips, and that map can be shared across an entire robot fleet.

Q: What can the F712 actually carry, and how fast does it move?
A: Up to 2,000 kg (4,409 lb) to heights of 8.5 meters (27.8 feet), with a top loaded speed of 1.7 m/s (3.8 mph) and positional accuracy of ±10 mm.

Q: Is it available now?
A: Yes — ABB says the Flexley Stack F712 is available immediately alongside its existing Flexley Tug and Flexley Mover AMRs.


Reference: "ABB Robotics includes vSLAM navigation in F712 autonomous forklift," The Robot Report. Published July 7, 2026. https://www.therobotreport.com/abb-robotics-includes-vslam-navigation-f712-autonomous-forklift/

Word count: 792 (4 paragraphs, 3 <h3> subtitles each preceded by <hr>, 1 <quote> block, Q&A included).

Note: I checked all 15 Tier 1 sources plus European Tier 2/3 sources (euRobotics, Fraunhofer, TU Munich, Oxford, ETH Zurich, UK AI Safety Institute) for articles dated exactly today — none had same-day posts. The ABB story was published July 7, 20:24 UTC, which lands in the early hours of July 8 in your local Jakarta timezone, making it the most current European Physical AI story available.