Inspekto

The problem: computer vision built from scratch every time

Inspekto was founded in Israel in 2017 by Harel Boren, who served as CEO, and his partner Yonatan Hyatt. The company operated from Tel Aviv and from another office in Heilbronn, Germany, close to the heart of the European car industry. The problem it chose was familiar to every quality manager in a factory: to set up a camera that inspects products on a production line, you call an integrator. The integrator chooses a camera, lens, lighting and software from thousands of options, and builds a solution tailored to one single point on the line. In a 2019 interview Boren described the process as tedious, expensive and dependent on professional services, and the result as a solution that cannot be moved elsewhere without being rebuilt.

The product: a system that learns from good examples

The product, INSPEKTO S70, was first shown in November 2018 at the Vision trade fair in Stuttgart. It put the software and all the vision components (camera, lens and lighting) in one box. The system trains on 20 to 30 images of good products and does not need examples of defects to start working: it learns what a good product looks like and flags anything that differs from it. According to a description published in Vision Systems Design, the software is built from three AI engines: one adjusts the optics for a sharp image, the second identifies the part, and the third performs the inspection. The company claimed that a quality manager could install the system on their own in 30 to 45 minutes. Boren himself qualified its marketing slogan, installation a thousand times faster at a tenth of the price, in the same interview, saying it was not literally accurate.

Customers and the Siemens acquisition

The early customers were car makers and their suppliers in Germany. In 2019 Boren listed Mahle, Bosch, Siemens, Daimler and BSH, and BMW ran the system in one of its plants. Siemens itself was a customer: it placed the system in a plant in Karlsruhe to check that all the components of a certain product had been assembled correctly before packing. In February 2024 Siemens bought Inspekto GmbH, the company's German entity, from the Heilbronn venture capital fund D11Z.Ventures. The deal amount was not published in the sources checked. The head of Siemens's factory automation division presented the acquisition as an addition to its portfolio of industrial AI software.

Under Siemens: a product line in industrial automation

After the acquisition the product was sold as Siemens Inspekto, with a Siemens industrial computer inside the kit. According to Siemens, it performs more than a hundred inspections a minute, trains on only 20 good examples, and connects to the factory's controllers and production management systems. It is important to understand the division of labor: the system does not decide by itself to stop the line. It detects a defect and sends an alert to the automation system, which acts according to rules set in advance. In April 2026 Horse Powertrain, an engine maker owned by Renault, Geely and Aramco, installed the system at its engine plant in Sweden, alongside a collaborative robotic arm. According to the company, a team can connect the system, test it on real products and decide whether to run it within two hours.

What this story teaches

The idea at the center of Inspekto is simple: training on good examples only. Defects are rare and varied, so it is hard to collect enough examples of them, while good products are available in abundance. This is what made it possible to move the setup from an integrator to a factory worker. However, almost all the known performance figures, such as installation speed and inspections per minute, come from the company itself or from Siemens, not from an independent test. The story also shows a familiar pattern in the Israeli industry: development in Israel, a first market in Europe, and a large industrial company that eventually buys a firm that was already its supplier.