Apera AI updates Apera Forge design and AI training studio

De-rackign simulation and training in Apera AI's Apera Forge.

De-racking simulation and training in Apera Forge. | Source: Apera AI

Apera AI Inc. this week announced updates to its Apera Forge, a web-based design and AI training studio that simplifies 4D vision-guided robotic projects. The latest release introduces new features to support advanced robotic cell design, end-of-arm-tooling (EOAT)-mounted camera configurations, and full simulation and training for de-racking applications.

“We give industrial robots 20/20 vision to unlock new levels of reliability and productivity in manufacturing,” stated Sina Afrooze, founder and CEO of Apera AI. “This release marks a new phase for Forge.”

“We’re not only expanding simulation realism and flexibility for advance cell design—we’re also empowering automation teams to train robust AI models for de-racking and EOAT-mounted vision setups, entirely on their own,” he added. “It’s another step forward in making advanced robotic automation faster, more accessible, and predictable.”

Apera Forge is a browser-based, no-code simulation and artificial intelligence training platform. The Vancouver, B.C.-based company said it enables manufacturers and integrators to build and validate vision-guided robotic applications, without the need for upfront investment in hardware.

By simulating robot, gripper, camera, part geometry, and cell environment within minutes, Apera AI said Forge reduces what traditionally takes weeks or months to mere hours.

Trained AI models can achieve greater than 99.9% reliability in object recognition and task performance. The company said it can deliver complete vision programs in 24 to 48 hours, ready for on-site deployment.

Apera AI advances cell design capabilities and more

The latest Forge release includes advanced cell design capabilities, explained Apera AI. The company said Forge now supports advanced cell design with greater flexibility, tuning camera placement, bin positioning, and multiple obstacles to better match real-world layouts and improve simulation results.

Users can now arbitrarily position cameras and bins and use reference cell CAD files to achieve more accurate cell visualization and scaling. Apera Forge also now supports integrated Obstacle Autopilot for enhanced robot navigation and collision avoidance in simulations. Users can import complete Apera Vue cell configurations, including bins, search regions, obstacles, tables, and calibrated cameras.

Apera Forge also supports EOAT-mounted, or “Eye-in-Hand,” vision for de-racking. This simplifies the design process by allowing users to preview and refine camera positioning directly in Forge. It can streamline design and validation for systems where cameras are mounted directly on end effectors, said Apera AI. Users can visually determine ideal camera placement by previewing camera views and ensuring clear, unobstructed vision of parts.

In addition, Forge now supports full simulation and AI training for de-racking applications, where consistent spacing, orientation, and rack structure are critical to effective part recognition and picking. Users can now precisely specify de-racking parameters, including initial positions, rack axes, part spacing, and part count.

Training robust AI vision models in Forge without Apera engineer support can be a significant time-saver for integrators and industrial manufacturers, the company claimed.

“De-racking is a highly common application in the automotive sector,” said Jamie Westell, director of engineering at Apera AI. “With our AI-powered 4D Vision deployed at the top 6 automotive OEMs in North America, this Forge release empowers their maintenance engineering managers to rapidly deploy vision-guided robotic automation across their plant for de-racking vehicle hoods, doors, body panels, and other racked parts.”

Apera AI said it aims to make factories more productive by empowering robots with human-like vision and advanced AI, dramatically increasing speed and accuracy. The company asserted that its sftware can handle challenging applications such as bin picking, sorting, packaging, de-racking, and assembly.