A Raspberry Pi and a camera running a vision model with no cloud
Brian Roemmele showed a Raspberry Pi with an attached camera running the open LLaVA 1.5 model locally to describe and recognize what passed in front of it.
Evidence: Hobbyist demo shared on X (2023). Not a factory deployment. Source date: 2023-11-30 · Checked 2026-10-23 · Tech: Raspberry Pi, camera, open-source vision-language model (LLaVA 1.5)
The problem
Camera-based checks used to mean expensive industrial vision systems and an integrator. Small shops couldn't justify that for simple questions like "is the part there?" or "is the bin empty?"
What they did
In a quick demo (in a building lobby, not a shop), he ran an open-source vision-language model on a Raspberry Pi with a camera, entirely on the device, and showed it identifying the people passing by.
Reported results
No production numbers. The point of the demo was that a credit-card-sized computer could run a useful vision model locally, without sending images to the cloud.
Our take
Since 2023 this has become much cheaper and easier: Raspberry Pi now sells a $70 camera with an AI chip built into the sensor, and a $70-$110 accelerator board. For a shop, the realistic first projects are simple: counting parts off a machine, confirming a part is seated before cycle start, or spotting an empty material rack. Keeping images on-site is also a plus for customers with strict data rules. Don't expect it to replace a CMM or a trained inspector.
Try this on Monday: Pick one yes/no visual question you answer by walking over to a machine several times a day. That's your pilot.
Brian Roemmele reports wiring an AI agent into an existing flexographic carton-box printing line, with small local models watching the machine and cameras feeding it visual data.
Evidence: Self-reported by the builder on X. Early, no independent verification yet. · via @BrianRoemmele
Neo Industries, a small machine shop in Colleyville, Texas, used automatic production monitoring to see how much of each cycle every tool was responsible for, then optimized tool by tool.
Evidence: Vendor-published customer story with named shop owner.
MSP Manufacturing in Bloomington, Indiana, uses an AI CAM plug-in to draft most of each program, with experienced programmers finishing and proving out the first run.
Evidence: Vendor-published Q&A with named company president.