n TECHNOLOGY September 2026 www.drivesncontrols.com 18 COGNEX HAS DEVELOPED a collaborative AI tool to simplify and scale AI-powered inspection in manufacturing. Since its beta launch in June 2025, more than 100 companies worldwide have been using the OneVision tool to accelerate their AI-powered vision deployment, with many moving from single-line applications to multi-site rollouts in days instead of months. According to Cognex, this reflects a broader shift as manufacturers move from isolated AI pilots to connected, enterprise-wide inspection strategies. OneVision is designed to address the challenge of how to deploy advanced vision applications at enterprise scale without adding complexity or slowing production. It uses a cloud-to-edge architecture, where AI models are trained, managed, and governed in the cloud, while inspection runs at the edge on Cognex vision systems. Users can manage the entire AI lifecycle centrally – from collecting and labelling production images, to refining models – and deploy updates across global fleets of devices. OneVision is designed to work with Cognex’s latest vision systems. By centralising model development and management, the tool will help to: n standardise inspection processes across sites; n reduce duplication of work across teams; n cut scaling costs by up to 50%; and. n maintain version control and consistency across deployments. “AI vision has long delivered value, but scaling it across operations has remained a barrier,” explains Cognex president and CEO, Matt Moschner. “Manufacturers encounter recurring challenges – from fragmented workflows to models that don’t adapt across environments. OneVision addresses this by unifying the simplicity of the edge with the scalability of the cloud, helping organisations to move from isolated pilots to consistent, enterprise-wide deployment.” “While OneVision leverages the cloud for development and management, runtime inspection remains fully edge-based,” adds Cognex’s vice-president of vision engineering, Reto Wyss. “Once a model is deployed, no connectivity to the cloud is required. Production images stay local and latency is not an issue.” www.cognex.com/en/products/ machine-vision-software/one-vision p The Californian energy conversion chip manufacturer Power Integrations has announced a GaN (gallium-nitride) technology rated at up to 2.2kV – higher than any other GaN technology. The PowiGaN technology will allow higher power densities and efficiencies, as well as safer, simpler architectures for applications such as industrial power conversion, energy storage, data centres, EVs, photovoltaic inverters and HVDC. The development extends the reach of GaN into voltage ranges traditionally served by SiC (silicon carbide) devices, offering an alternative high-frequency technology. p The Israeli servodrive specialist Elmo Motion Control has formed a partnership with Air, the US developer of advanced aircraft, to integrate high-performance motion controls into Air’s Heavy Lift UAS – an uncrewed VTOL (vertical take-off and landing) cargo aircraft capable of carrying 250kg . They will combine Elmo’s high-power-density servodrives with Air’s next-generation electric aviation designs to improve aircraft efficiencies and performance. Air-cooled Elmo Gold HV drives will power eight electric propulsion motors, delivering precise, responsive control. p Neura Robotics, the German cognitive robotics developer, has formed a partnership with RWTH Aachen University to establish a Neura “gym”. It is part of a global network of gyms for training and deploying physical AI, designed to tackle the scarcity of real-world training data. Ten gyms are being developed, with the first five, in Europe, the US and China, due to be running by end of 2026. Unlike large language models, which train on trillions of data points, robots have access to a fraction of that data. Simulations alone cannot reproduce the variability and unpredictability of the realworld environment that robots operate in. At Neura’s gyms, all types of cognitive robot will be able to train under real-world conditions to generate data for the a cloud-based platform. p The Italian industrial computing developer, Sfera Labs, has announced a compact DIN-railmounting PLC based on the Raspberry Pi 5 computer. The Iono Pi v3 is aimed at engineers who need the Pi 5's processing power in a ready-to-deploy format with many analogue and digital I/O lines, including 16-bit precision analogue inputs. The PLC supports the Raspberry Pi 4B and Pi 5 computers in memory configurations from 1–16GB, with Gigabit Ethernet, Wi-Fi, Bluetooth/BLE and USB communications. https://sferalabs.cc/product/iono-pi-v3 p Toshiba Electronic Devices & Storage has started mass production of compact microcontrollers with Arm Cortex‑M4 cores and FPUs (floating‑point units) for next-generation industrial equipment and consumer products. The TXZ+ M4L controllers, aimed at singlemotor applications, integrate motor-control hardware, security functions, and functional safety on one chip. By offloading motor-control processing to dedicated hardware, they reduce the CPU load, enabling precise, efficient motor control. TECHNOLOGY BRIEFS Time-saving AI tool helps to scale vision tasks FANUC AND REGAL REXNORD have announced a strategic collaboration to combine Fanuc cobot (collaborative robot) systems with Thomson’s linear axes to produce out-of-the-box systems that extend cobot operating ranges up to 10 metres. They say that a jointly developed plug-in interface between Fanuc’s CRX cobots and Thomson’s Movotrak CTU (cobot transfer unit) linear axes will unlock “new levels of collaboration between humans and robotic systems”. The CTU is an external linear axis designed to extend cobot operating ranges. It can support Fanuc’s largest cobots, including its CRX30iA model, which has a payload of 30kg and a reach of 1756mm. The “seventh axis” can move the cobot from one workstation to another along a rail that is fixed to the floor, wall or ceiling. The axis is claimed to be the first of its type to offer collision detection that stops movement of the axis and cobot if it encounters an obstacle. Thomson and Fanuc say that integrating linear motion, gearing, couplings, motors and software into a single system will simplify the implementation of advanced robotics, while improving performance, reliability and speed to deployment. “Adding an external axis transport layer to a 6-axis cobot opens new avenues for productivity, innovation and profitability for our customers,” explains Emily Kern, president of Regal Rexnord’s Linear Motion Division. Fanuc and Thomson join forces to extend cobot operating ranges up to 10m Pilot users of Cognex’s OneVision AI tool have reported faster deployment, improved throughput, and new levels of collaboration
RkJQdWJsaXNoZXIy MjQ0NzM=