Overview AI launches the OV Spark line of AI inspection cameras

Overview AI today launched the OV Spark and OV Spark Pro. The new cameras constitute a new line of AI inspection cameras with a level of artificial intelligence that answers questions in plain English.

Spark comes with 10 inspection tools that can be mixed and matched into a single recipe. Seven are deep learning AI models: AI Trigger, Alignment, AI Classification, AI Segmentation, AI Counting, AI OCR, and Anomaly Detection.

The other three are inspection ready, with the ability to add images when an inspection requires training. This includes barcoding, measurements and color matching.

“In a robot cell, the part doesn’t always land the same way twice,” said Christine Chacko, technical product marketing manager at Overview AI The robot report. “OV Spark anchors itself to the characteristics of the part itself, so each inspection area follows it as it moves or rotates.”

“If no part is found, the capture fails rather than passing a blank frame,” he explained. “It contains 10 inspection tools, and the PLC can select the recipe by ID as the arm moves between stations. Adding a station therefore only requires a new recipe number, not another camera.”

The recipe, model training, inference, judgment, PLC communication, image library, and operator interface all continue to run on the camera itself.

“We’ve been introducing AI to factories since 2019, and the hardest part isn’t the inspection itself. It’s that every new line means another project, another specialist, more weeks,” said Chris Van Dyke, co-founder and CEO of Overview AI. “Our mission is to find every defect in every factory, so the camera has to do the setup work itself. Spark is our best product that can do exactly that.”

Overview AI said that OV Spark and OV Spark Pro are now available directly from the company and through its distribution partners.

Why did Overview AI create Sparky?

Despite improvements to the accuracy of its vision models, Overview AI said customers were still seeing a gap in consistency. Which tool to use and what to extract from the production data still depended on individual judgement. This led Overview AI to create Sparky, the AI ​​agent that lives on the Spark camera.

Sparky gives teams a shared starting point for these judgments, the company said. This allows all members of a team to equally understand the reasoning behind an inspection.

“The onboard officer tells the operator which instrument needs a check and why. Then Sparky answers questions about any catch in plain English,” Chacko said. “A robotic cell can produce thousands of acquisitions that no one ever examines. An operator can ask Sparky to understand what failed in a certain period of time and receive data that helps preserve the model for the next set of inspections.”

When it comes to agent tool selection, Sparky recommends tools for an inspection and explains the reasoning behind each. When asked about a connector that needs to be confirmed to have 24 pins before being sent to a PLC, Sparky explains that the alignment tool is needed to follow the part as it moves and the AI ​​Counting tool is needed to confirm that every pin is accounted for.

For agent diagnostics, Sparky makes the capture library searchable in plain language. This means that a question about the previous round returns a timeline of acquisitions, each with its verdict and associated tool results.

As a human-machine interface (HMI) builder, Sparky prepares the operator’s screen from a description of the data a station needs, using widgets created to show product metrics at a glance.

“For years, manufacturers have told us the same thing: They have more inspection points than qualified people to install them. Spark is our answer to that question,” said Russell Nibbelink, co-founder and chief operating officer of Overview AI. “A quality engineer who knows the part but has never trained a model can now perform an inspection on the line, and that changes who in a plant can actually do this work.”

OV Spark presents new hardware

The 1.2MP OV Spark and 5MP OV Spark Pro share the same sealed, all-in-one body, which is IP66 rated and small enough (70 x 116 x 53mm) to fit into existing cells. Spark can be powered over Ethernet via a PoE coupler.

“We took seven years of what we learned implementing AI inspection in real plants and compressed it into a single camera: Spark,” said Xiaoyang Kao, co-founder and chief technology officer of Overview AI. “We designed it on Qualcomm Dragonwing and put the entire agent stack into the device. The camera recommends tools, reads production data and presents the operator’s screen itself. We have always run intelligence at the forefront. Spark is the first product where such intelligence can support an expert in doing their job.”

Mansfield Engineered Components uses OV Spark

The OV Spark is already in operation at Mansfield Engineered Components (MEC), a metal fabricator and contract manufacturer in Mansfield, Ohio. There, the camera sits on an end-of-line station that has not previously undergone any formal inspection, but only a visual inspection by the operator.

A missing washer is the most common defect on that line, and the inspection was performed by a quality engineer who had never worked with a vision camera before joining the company a year ago. MEC has since moved the camera between two lines and is examining a second Spark for a check of the rivet curl and stamped date code.

“I’m a quality engineer with a degree in industrial engineering, I’ve been working for about a year, and I had minimal experience with cameras before,” said Logan Hoon, the quality engineer at MEC who carried out the inspection, of what it took to get the station up and running.

“The part I expected to be difficult was the aligner, and it turned out to be pretty easy, because [the camera] it tells you exactly what to do,” Hoon said. “I think we knocked it out of the park. [Spark is] reliable, easy to use and easy to train for maintenance workers.”