Asset condition assessment streamlined.

Thermal imaging cameras to bring AI-driven asset condition assessment directly onto the camera.

Flir has announced the worldwide availability of the MentorLens® app by MentorAPM on its professional-grade iXX-Series thermal imaging cameras, bringing AI-driven asset condition assessment directly onto the camera. Developed by MentorAPM, a provider of asset performance management software for critical infrastructure, the app allows maintenance and reliability teams to capture nameplate, condition and thermal imagery on a single device and send it straight to the asset record, without the file transfers and manual data entry that have long slowed condition monitoring programs.*

In industrial operations, the value of a thermal inspection is determined less by the image itself than by what happens to it afterward. Files are moved from an SD card to a laptop, thermal metadata is altered or lost along the way, and findings end up in a folder rather than in the system where maintenance work is actually planned. Running natively on the iXX-Series, the MentorLens app removes that handoff. Images upload from the camera directly to the asset record with their thermal data intact, and technicians can create new assets or update existing ones without leaving the camera. All Flir imaging and hardware controls remain available, including MSX, palette selection, lamp and laser, so the same device that captures the thermogram also captures the nameplate photo and the condition photo.

Once the images are synced, MentorLens processes them automatically. Nameplate and tag data is transcribed into standard asset attributes, visual imagery is scored for condition, and thermal imagery is measured against operating baselines taken from the asset's own nameplate data or from manufacturer manuals. Reports return observed anomalies, temperature readings and recommended actions, and because previous assessments for the same asset are carried into the analysis, degradation is tracked across inspections rather than judged one image at a time.

“Customers don't need more inspection data. They need actionable information that helps them prevent downtime and make better maintenance decisions," said John Gould, Sr. Director of Strategic Business Development, FLIR. "With MentorAPM integrated into the iXX-Series workflow, thermal findings become part of a broader asset health picture. Data is automatically evaluated against previous conditions and delivered to the teams responsible for maintenance planning, enabling earlier intervention and more effective predictive maintenance programs.”

For the wider organisation, the benefits go beyond the inspection itself. Condition and thermal results feed the MentorLens platform, where rules can automatically trigger preventive work orders and where condition history, work history and SCADA events are held against the same asset record. Reports are exported as standardised PDFs, and the underlying data can be pushed by API into an existing CMMS or EAM system, so operators are not obliged to change their system of record in order to benefit. Every AI-assisted workflow keeps a human in the loop, with a person confirming what is committed against the asset.

“For most operators the hardest part of asset management is still the most basic one, which is knowing what you own and what condition it is in,” added Tacoma Zach, CEO of MentorAPM. “Putting our app on the iXX-Series means high-quality thermal data now enters that picture at the point of capture, from the same technician and the same device. In plants where downtime is measured in millions of dollars an hour, the payback on that is very quick.”

First proven in water and wastewater utilities, the combined solution is now being adopted in manufacturing, metals, and oil and gas, where thermographic inspection has typically been outsourced to specialist contractors at considerable expense. MentorAPM reports that customers using MentorLens achieve up to 10x faster field data collection and cost savings of more than 75 percent against conventional condition assessment methods. The MentorLens app is available worldwide now for Flir iXX-Series users with a MentorAPM subscription.

---

* Both companies will present the integration at the SMRP 34th Annual Conference, held 28 September to 1 October 2026 at the Raleigh Convention Center in Raleigh (NC USA). Tacoma Zach, CEO of MentorAPM, leads an Innovation Lab session titled "From Camera to Asset Intelligence: AI Workflows for Reliability Teams" on Tuesday 29 September, 1:15 to 2:15 PM ET, in Innovation Lab #3. The session walks through practical workflows using MentorLens with Flir thermal cameras to capture equipment information, analyse visual and thermal conditions and generate structured asset intelligence automatically. John Gould of Teledyne FLIR joins the session to discuss the iXX-Series.

@flir @MentorAPM  @mepaxIntPR  #PAuto #Manufacturing

Shutter sensors for machine vision.

Nexora™ is a new family of CMOS image sensors from Teledyne e2v, that delivers high sensitivity, high-speed imaging and low power consumption for demanding machine vision applications.

Rafael Romay, President of Teledyne Vision Solutions, said: “With Nexora, we focused on giving camera manufacturers a new platform that balances image quality, speed and integration simplicity. The result is a sensor family that can support a wide range of current and emerging machine vision applications.”

Available in 12 MP (4,096 by 3,072) and 16 MP (4,096 by 4,096) resolutions, in both monochrome and color versions, Nexora is designed for applications that require sharp images at very high speeds. These include high-resolution machine vision, intelligent traffic systems, and outdoor surveillance.

At the heart of Nexora is Teledyne e2v’s proprietary backside-illuminated (BSI) pixel technology. By positioning the photodiode closer to the incoming light and maximizing the light-sensitive area of each pixel, the sensor achieves higher light-collection efficiency than conventional front-side illuminated (FSI) architectures. This results in quantum efficiency (QE) of up to 73% at 550 nm and 15% at 940 nm, improving performance in challenging lighting conditions and near-infrared applications.

The Nexora family also features Teledyne e2v’s 3D-stacked sensor architecture, which integrates multiple functions within a compact, multilayer structure. This approach increases design flexibility while helping camera manufacturers develop smaller, higher-performance imaging systems.

It incorporates a versatile LVDS interface that is compatible with a wide range of FPGAs and image signal processors. It is housed in a compact 23 by 21.5 mm package that is compatible with 29 by 29 mm² cameras and standard C-mount optics.

---

 @e2vteledyne #PAuto  #machine_Vision

Machine vision trends.

Five trends set to define industrial machine vision in 2027.

Five trends set to shape machine vision as the industry moves into 2027 have been outlined by BitFlow. The company also detailed how its flagship CoaXPress (CXP) and Camera Link solutions are positioned to power the next generation of Ai-optimised inspection systems. These systems span critical sectors including semiconductors, EV battery manufacturing, automotive, pharmaceuticals, and food processing.

CXP4

The outlook follows a milestone year for the company: the full production release of BitFlow's Claxon CXP-12 lineup — five models spanning the single-link Claxon CXP1 through the quad-link Claxon CXP4 and long-reach Claxon Fiber — purpose-built to feed NVIDIA GPU-accelerated inference pipelines at full CoaXPress 2.0 throughput. With that foundation now in full deployment, BitFlow is turning its attention to the next wave of standards and workloads set to arrive in 2027.

"We're watching a fundamental shift in how manufacturers approach quality control, and it's only accelerating as we head into next year," said Donal Waide, Director of Business Development, iSystems, Advantech. "Traditional rule-based vision cannot deliver the defects-per-million rates modern production demands. BitFlow frame grabbers provide the zero-latency, zero-frame-loss image acquisition infrastructure that will make Industrial Vision Ai the operating standard in 2027."

1. Industrial Vision Ai Becomes the New Quality Standard: As manufacturers push toward near-zero defect rates going into 2027, deep learning and adaptive Ai models are set to replace conventional inspection methods at scale. BitFlow's ultra-reliable, high-bandwidth architecture feeds massive multi-camera, high-resolution image streams directly to Ai accelerators without bottlenecks. This is a critical capability for automated optical inspection (AOI) in electronics and precision manufacturing. Advantech Vision Ai deployments already show real-time defect detection meeting the world's strictest quality standards in 2026, and adoption is on track to accelerate further next year.

2. CoaXPress v3.0 Arrives to Reset the Performance Bar: CoaXPress 2.0 remains today's performance leader, delivering up to 12.5 Gbps per link (CXP-12) with power, triggering, and control over a single coaxial cable. Link aggregation scales to 50 Gbps across four links and beyond. BitFlow offers one of the industry's most comprehensive CXP-12 portfolios, from compact low-profile designs to multi-link systems supporting aggregate bandwidths exceeding 100 Gbps.

But the headline for 2027 is CoaXPress v3.0. The CoaXPress technical committee has been finalising the new specification since late 2025, with release still expected before year-end 2026 though an official date has not yet been announced. The headline numbers are already locked in: a nominal 25 Gbps per link (CXP-25), full integration of CoaXPress over Fiber into the core standard rather than an add-on, an expanded uplink for faster triggering, and support for forward error correction. That means systems designed today should be built with a clear runway to next-generation speeds, not a forklift upgrade next year.

BitFlow is already there: its Claxon Fiber series and QFi copper-to-fiber converter use QSFP+ transceivers to deliver long-distance transmission over hundreds of meters and complete EMI immunity today, on an architecture built to carry forward into v3.0 fiber-native deployments. "We're already positioning products to support next-generation CoaXPress v3.0 speeds and fiber-native architectures," Waide noted. "Customers building on our platform now will be ready the day v3.0 hardware ships in 2027." 

3. Edge Ai Demands Sub-5 Microsecond Latency: In high-speed production lines, latency directly impacts throughput. BitFlow's edge-native architecture eliminates bottlenecks through:

• Direct GPU transfers (GPUDirect RDMA for NVIDIA, DirectGMA for AMD) bypassing the CPU for sub-5 µs host-to-GPU latency
• FPGA-based pre-processing pipelines executing debayering, flat-field correction, LUTs, and ROI extraction at full pixel clock rates
• Up to 128 independent DMA streams for complex multi-camera and region-of-interest configurations

This deterministic, high-speed acquisition lets Ai models focus purely on inference. This capability is driving explosive growth in inline EV battery inspection and high-speed sorting applications, and one BitFlow has already demonstrated in production: its Claxon Fiber-over-CoaXPress frame grabber has been integrated with an Advantech Ai inference edge computer and NVIDIA TensorRT for real-time pose-estimation workloads, underscoring the same low-latency pipeline now being applied to industrial inspection.

4. 3D Vision and Robotics Integration Accelerates:  Demand for precision 3D inspection using stereo vision, time-of-flight, and laser triangulation is surging in battery electrode metrology, robotic bin-picking, and assembly verification. It is projected to intensify further in 2027. BitFlow's high-bandwidth CXP interfaces deliver synchronised multi-view data flows essential for 3D reconstruction and Ai pose-estimation algorithms. This directly supports the collaborative robot (cobot) and autonomous mobile robot (AMR) boom, where real-time object recognition and human-robot collaboration require rock-solid image pipelines for Industry 5.0-style manufacturing.

5. Extreme Resolution Meets Deterministic Performance: As cameras push toward higher resolutions, faster frame rates, and event-based sensors, BitFlow's latest solutions, including the QFi for remote CXP deployments, are built to handle extreme specifications without compromise heading into 2027, supporting long cable runs, precision encoder-based triggering, and error-free transmission in harsh industrial environments.

The BitFlow-Advantech Advantage. "As part of Advantech, we deliver complete end-to-end solutions: world-class frame grabbers integrated with powerful edge servers, AI software stacks, and modular architectures," Waide stated. "Since 1993, BitFlow's focus has remained constant: providing deterministic, high-throughput image acquisition that makes advanced vision possible."

Whether manufacturers are scaling Vision AI for defect-free production, transitioning to fiber for extended reach, or powering autonomous inspection systems, BitFlow frame grabbers deliver reliable performance, every frame, every time.

"As companies become more familiar with the advantages of AI based systems, it's always good to know that Advantech is an Elite partner of NVIDIA and have been planning this technology breakthrough for some time now. When the customer is ready, we will be waiting," cites Waide.

---

@Advantech_IIoT @AdvantechIA  @OConnell_PR #Bitflow #PAuto #Ai

Camera application expanded.

Camera data goes directly to PC, eliminating CPU overhead and reducing latency.

SVS-Vistek has now expanded its FXO platform by introducing the new FXO 25GigE Series of machine vision cameras enabled with Remote Direct Memory Access (RDMA) technology*.

Combined with the camera's 25GigE interface, the integration of RMDA greatly reduces latency and response times to achieve near real-time image processing. In addition, the use of cost-effective SFP28 optical transceivers -- rather than traditional RJ45 connectors -- allows data transmissions over cable distances of up to 10,000 meters.

Wide range of models.
Built on 3rd and 4th generation Sony® IMX global shutter sensors, SVS-Vistek FXO 25GigE Series cameras are available in color and monochrome configurations offering resolutions and frames-per-second (fps) of:

• 1.8 megapixel (1600 x 1140 pixels) x 671 fps
• 5 megapixel (2448 x 2048pixels) x 262 fps
• 8.1 megapixel (2848 x 2848 pixels) x195 fps
• 12.3 megapixel (4096 x 3008 pixels) x182 fps
• 16.2 megapixel (5328 x 3040 pixels) x 144 fps
• 20.4 megapixel (4512 x 4512 pixels) x109 fps
• 24.6 megapixel (5328 x 4608pixels) x 96 fps.

This range of options makes the FXO 25GigE Series ideal for diverse applications requiring uninterrupted, stable data flow of extremely high resolutions and frame rates.

Regardless of the FXO 25GigE model, the cameras integrate seamlessly into current network structures, enabling end-users to adapt them without expensive hardware overhauls. The GigE Vision standard ensures compatibility with existing machine vision software and systems, while GenICam 3.0 compliance allows for plug-and-play functionality, reducing complexity.

Rounding out the FS 25GigE's impressive feature set are LUT, binning, ROI, electrical and optical inputs, a 4-channel strobe controller, and an industrial TTL-24V I/O interface equipped with SafeTrigger, logic functions, programmable sequencers and timers, plus an RS232 interface. The milled housing measures 50 x 50 x 100 mm, making it currently one of the most compact industrial cameras with a 25GigE interface in the world. The housing's temperature management also provides excellent image homogeneity.

---

* RMDA works by transferring image data from FXO 25GigE cameras directly to the main memory of a host PC, reducing stress on the operating system and allowing the CPU to perform other tasks by freeing up computational resources. Consequently, CPUs and GPUs can focus on intensive functions like image processing, feature extraction, or neural network computations rather than managing network I/O. In machine vision, where real-time processing of high-resolution images or video streams is critical, RMDA ensures faster decision-making. A firmware update also lets the camera be switched to conventional streaming via GVSP (GigE Vision Stream Protocol).

@OConnell_PR #svs_vistek #PAuto

Machine vision camera.

The launch of an ultra-high resolution industrial camera that merges the power of a Sony 245.8-megapixel CMOS area sensor with the lightning-fast CoaXPress-12 (CXP-12) quad interface hs been announced by SVS-Vistek. One of the most advanced cameras ever introduced, the new SVS-Vistek shr811CCX12 enables the acquisition of precise, detail-rich 19200 x 12800 pixel color images at data transfer speeds of 12.4 frames per second to significantly improve quality, reliability, and yield.

The shr811CCX12 heightens accuracy and propels throughput in a variety of applications, especially those requiring a large field of view or analysis of intricate patterns. It brings unprecedented performance to surface inspection of semiconductor wafers, PCBs, solar panels, and high pixel density FPDs, while also having the capabilities to tackle tasks such as medical imaging, astronomy surveys, city mapping, aerial and ground surveillance, and Intelligent Transportation Systems.

The camera harnesses the next-generation CXP-12 interface that simplifies application design and setup with the familiar, easy-to-use CoaXPress standard. The shr811CCX12 features four CXP-12 micro-BNC connections, each capable of a maximum data transfer rate of 12.5 gigabits per second (Gbps) and 24VDC Power over CoaXPress (PoCXP). When these four links are aggregatned, the camera will achieve rates of 50 Gbps.

The latest model in SVS-Vistek's industrial proven SHR camera platform, the shr811CCX12 shares the SHR's high dynamic range, boosting light sensitivity with minimal noise. With more pixels capturing light, the camera can distinguish subtle differences in color or texture that low resolution cameras miss, improving performance in challenging lighting conditions. Like other SHR cameras, the shr811CCX12 also comes equipped with a comprehensive I/O with signal voltage, RS232, and galvanically isolated input, along with a sequencer, binning, and an integrated multichannel LED strobe controller.

The shr811CCX12 is engineered to withstand harsh industrial environments which would otherwise adversely affect image quality. It is protectively housed in an IP30 thermally-optimized metal enclosure that features a large M72 mount that can be easily adapted to any lens, and provides an operational temperature range of -10° to 60° C (14° to 140° F).

---

#SVS_Vistek @OConnellPR #PAuto

Acquisition bolsters machine vision capability.

The Headwall® Group (“Headwall”) has acquired EVK DI Kerschhaggl GmbH (“EVK”), an Austria-based technology company specializing in industrial sensor-based sorting and inspection systems. This strategic acquisition marks a significant step forward in Headwall Group’s mission to enhance its portfolio and bolster its presence in the industrial machine vision market.

Mark Willingham, CEO of Headwall (left),
and Matjaz Novak, CEO of EVK 

The integration of EVK into the Headwall Group is a natural extension of Headwall Group’s strategy to diversify its product offerings and deepen its capabilities in delivering comprehensive, high-value imaging solutions. EVK’s innovative hyperspectral and inductive sensor technologies as well as data analysis expertise align seamlessly with Headwall Group’s existing products and commitment to advancing hyperspectral imaging applications and AI-driven interpretation software in machine vision and remote sensing markets.

“We are thrilled to welcome EVK into the Headwall family,” said Mark Willingham, CEO of Headwall Group. “Their proven track record in sensor-based solutions and industrial inspection complements our vision of providing integrated, next-generation hyperspectral technologies at the speed and scale needed by the industry. This partnership will drive further innovation and accelerate our ability to meet the increasing demand for more advanced, application-focused imaging solutions.”

Matjaz Novak, CEO of EVK, stated, “Joining forces with Headwall presents a tremendous opportunity to combine our expertise and technologies. Together, we can offer our customers enhanced solutions and expand our reach in the industrial imaging market. We look forward to a collaborative future that leverages the strengths of both organisations.”

EVK’s cutting-edge technology, which spans a range of industries including food processing, plastics recycling, and material sorting, will empower Headwall Group to broaden its market reach and foster new growth opportunities in both machine vision and remote sensing applications.

---

@headwall @__EVK__ #PAuto #Photonics

Event-based machine vision.

Camera enables ultra-fast, efficient machine vision with event-based sensing.

The IDS’ new uEye EVS camera line* incorporates the high-speed, dynamic range and data efficiency of the Prophesee-Sony IMX636HD event-based vision sensor to offer new capabilities for industrial machine vision applications. Thanks to the event-based sensor technology, the cameras only capture relevant changes in the image. 

The result of extensive collaboration between the two companies, the solution features Prophesee’s proven neuromorphic approach to capturing fast-moving objects with significantly less data processing, power and blur than traditional frame-based methods. With these capabilities, the uEye EVS camera is the ideal solution for applications that require real-time machine vision processing at very high speed, such as optical monitoring of vibrations or high-speed motion analyses.

The camera benefits from Prophesee’s event-based vision’s ability to capture only relevant events in a scene. In contrast to conventional image sensors, it does not capture every image completely at regular intervals (frames) but only reacts to changes within a scene. It transmits events depending on when and where the brightness in its field of view changes - for each individual sensor pixel. The temporal resolution, i.e. the minimum measurable time difference between two successive changes in brightness, can be less than 100 microseconds.

The sensor is supported by Metavision SDK, a seamlessly integrated suite of software tools and models, APIs, and other training and development resources from Prophesee for efficient evaluation, visualization, and customization.

"This partnership combines our mutual areas of expertise to realize the benefits of event-based vision, including remarkable temporal resolution which make the cameras optimised for analysing highly dynamic scenes. It enables best conditions for capturing fast object movements without loss of information, comparable to an image-based frame rate of more than 10,000 images per second," explains Patrick Schick, Product Owner 3D & Vision Software. “At the same time, the sensor ignores all motionless areas of its field of view and thus generates 10 to 1000 times less data than image-based variants. This saves memory and computing time.”

“IDS cameras are well known to address the toughest machine vision use cases and with the incorporation of Prophesee event-based vision technologies, it strengthens its offering to provide far more performance, power efficiency and accuracy, even in the most challenging conditions,” says Luca Verre, CEO and co-founder of Prophesee. “We are excited to see how the efforts of this tight collaboration have resulted in the new uEye EVS camera which leverages the potential of our sensors and development environment to deliver new value to its customers.”

---

*See also Camera has with event-based sensor. (1/3/2025) 

@IDS_Imaging  @mepaxIntPR #Pauto Manufacturing #Vision 

 

Tools for in-line 3D measurement and inspection.

The Z-Trak™ 3D Apps Studio, a suite of software tools developed for in-line 3D machine vision applications has been introduced. Designed to work with Teledyne DALSA’s Z-Trak family of laser profilers, it simplifies 3D measurement and inspection tasks on production lines. Capable of handling 3D scans of objects with varied surface types, sizes and geometric features, Z-Trak 3D Apps Studio is ideal for factory automation applications across a wide range of industries including electric vehicles (EV batteries, motor stators etc.), automotive, electronics, semiconductors, packaging, logistics, metal fabrication, lumber, and many more.

It features streamlined tools for measuring object thickness, inspecting glue-beads, weld seams, and identifying defects on flat, inclined, and curved surfaces on machined, assembled, or extruded parts. It also includes anchoring and data enhancement features like reflection elimination to ensure reliable and repeatable results under diverse operating conditions.

The Z-Trak family of laser profilers features various multi-sensor topologies to enhance the field of view while preserving height resolution, to overcome occlusion, or to provide a 360° view of objects for inspection and measurements. Z-Trak 3D Apps Studio simultaneously acquires, processes, and analyzes 3D scans and 2D gray scale images (reflectance data) all with 3D visualization capabilities.

For easy integration and quick deployment, the Z-Trak 3D Apps Studio is accessible via Sherlock 8.30 (or higher), a field proven, no-code, graphical development environment for factory floor applications. Z-Trak 3D laser profilers include a Sherlock 8 license for in-line measurement applications.

---

@TeledyneDALSA  @Notified #PAuto #Manufacturing

Prototyping of End-to-end AI Accelerated Vision Applications.

BitFlow has successfully integrated the unprecedented computing power of the NVIDIA® Jetson AGX Orin module, which can perform 275 trillion operations per second (TOPS), with the lightning-fast data rates of its CoaXPress (CXP) frame grabbers.

When combined with the NVIDIA AGX Orin Developer Kit, this cost-effective platform empowers engineers to prototype complex machine vision and autonomous inspection applications, leveraging AI accelerated image processing while simultaneously supporting up to four CoaXPress (CXP) cameras and multiple concurrent AI application pipelines. Groundbreaking new applications are more easily developed that augment rule-based machine vision with image-based analysis, making it possible to move beyond “pass/fail” to tasks like image classification, image segmentation, and object detection.

Once proof-of-concept is established, a production model can move forward utilizing an Advantech AIR-030 AI Inference System Box featuring PCI Express x16 and based on the NVIDIA Jetson AGX Orin. As a result of this innovation, Time to Market and associated development costs are significantly reduced.

50GB Data Transfer. BitFlow CXP frame grabbers connect directly to the Jetson AGX Orin via a built-in x16 PCIe slot. Image data may then be transferred at speeds up to 50GB per second from CXP cameras to the NVIDIA Ampere GPU architecture — much faster than what NVIDIA Jetson users are typically limited to using USB3 or GigE Vision cameras. BitFlow CXP frame grabbers DMA directly into the embedded GPU memory for image capture, pre-processing, and machine learning inference, shifting the load from the host computer to avoid CPU overhead.

Besides faster transfer speeds, the CoaXPress interface allows a single cable to carry all data, control, triggering, and up to 13W of power to connected cameras at lengths as far as 100 meters. CoaXPress eliminates the need for multiple cables and a local power supply, therefore giving the system integrator far more flexibility for their prototype designs.

Seamless integration between the NVIDIA Jetson AGX Orin and BitFlow frame grabbers is achieved through BitFlow’s Linux AArch64 SDK. With the SDK being universal, not only is BitFlow's full line of CXP frame grabbers (Cyton and Claxon families) supported, but additionally the BitFlow Axion Camera Link family is an option for customers.

---

@Advantech_IIoT @AdvantechIA #Bitflow #Manufacturing #PAuto 

Machine vision range.

Teledyne FLIR IIS (formerly Point Grey Research) has introduced the new Dragonfly® S USB3 camera series to its machine vision portfolio. The new series is a result of Teledyne’s constant commitment to set and enhance vision industry standards for customer success. It addresses the market’s essential need for a modular, compact, and lightweight camera for at-scale manufacturing, volume-based applications, and multi-camera systems.

Dragonfly S has a modular design to accelerate the inception stage of imaging application development for a diverse range of industrial and non-industrial segments. “This is a first-of-its-kind practical option for factory automation and embedded vision-based device manufacturers,” said Sadiq Panjwani, General Manager at Teledyne FLIR IIS. “The modular concept of the new camera empowers engineers to quickly test, develop, and deliver competitive vision products to market at scale.”

The camera series is ideal for embedded or handheld device applications such as biometrics kiosk solutions, ophthalmoscopy, 3D scanning, automated optical inspection, and more. It is available in multiple configurations, from board-level to a fully-enclosed module in a ruggedized aluminum case. The USB connector is offered in a rear or side orientation with screw-in locking mechanism. These modular options provide maximum flexibility and durability, allowing seamless integration in tight spaces. The cased models are Class B EMC safety compliant as mitigating electromagnetic interference is critical for healthcare and residential applications. Additionally, an onboard image buffer ensures that every captured image frame is reliably transferred to the host CPU.

The first of its series Dragonfly S is now available for pre-order.

---

@flir #PAuto #PointGreyResearch #Vision

Revolutionizing machine vision in factory automation applications.

Next generation camera platform and imaging software leverage years of machine vision GigE camera design expertise and customer understanding to deliver best-in-class imaging reliability, quality, and performance.

Advancements in industrial automation require machine vision technologies to work in extreme environments while ensuring the production capabilities are highly efficient. Teledyne FLIR IIS is bringing together Teledyne’s long-standing imaging expertise, deep factory automation knowledge, and a focus on innovation and customer success, to alter the factory automation market landscape and transform how machine vision is used.

“Teledyne FLIR IIS is committed to launching imaging technologies that will build the future of machine vision,” said Sadiq Panjwani, General Manager at Teledyne FLIR IIS. “We stay focused on enhancing customer experience to deliver best-in-class imaging reliability and performance – for both our industrial and non-industrial customers.”

The latest Teledyne FLIR IIS imaging product releases are architected to work under extreme conditions ensuring constant reliability without compromising imaging performance

Spinnaker™.
Spinnaker 4 is the next-generation GeniCam3 API library for machine vision developers to evaluate, test, and deploy products quickly with the highest performance and reliability. The new software release includes groundbreaking advancements in GigE camera performance and reliability, leveraging Teledyne’s decades of experience delivering best-in-class machine vision GigE cameras worldwide. Intense camera system performance testing, including operating imaging systems with 40 GigE cameras continuously for multiple days at a time, resulted in no incomplete images. Flawless image capture is maintained even when pushing the system CPU usage to over 90%, with intensive image processing and multiple background applications. The software enables fast camera detection and image processing, and much more. 

Forge™.
Teledyne FLIR IIS’s newest camera platform, Forge, utilizes Spinnaker’s new software architecture and is available in a variety of camera models across multiple sensors. Forge is designed to easily build robust and powerful imaging systems – faster. It offers flexible link speeds as well as the ability to go beyond 5GigE performance and control data transfer to the host. For more information, visit our website.

These latest imaging innovations from Teledyne FLIR IIS represent the start of real change in the factory automation market with new options for best-in-class imaging reliability, quality, and performance.

---

@flir #TandM #PAuto

Emerging sensor types for embedded vision.

by Dr Matthew Dyson, Principal Technology Analyst at IDTechEx.

Anyone who has used a ‘smart lens’ app on their smartphone is familiar with machine vision. Rather than simply reproducing and storing a picture for later viewing, machine vision applies an image processing algorithm to obtain additional insight, such as locating edges or object identification. While few would regard a ‘smart lens’ app as essential, machine vision is a critical component of many industrial processes, such as material sorting and quality control. It is also crucial for ADAS (advanced driver assistance systems) and, ultimately, autonomous vehicles.

Embedded vision brings these computational capabilities to the ‘edge’. Rather than images being sent from a sensor to a central processor, initial analysis is performed adjacent to the sensor on a dedicated, often application-specific, processor. This greatly reduces data transmission requirements since rather than sending all the acquired information (i.e., each pixel’s intensity over time), only the conclusions (e.g., object locations) are transmitted. The combination of reduced data transmission and application-specific processing also reduces latency, enabling quicker system responses.

	Comparing embedded and machine vision. Source: IDTechEx

	Minimizing Size, Weight, and Power (SWAP) While in many image sensing applications maximizing performance metrics such as resolution and dynamic range is the priority, sensors for embedded vision are typically designed with other priorities in mind. Collected data needs to be good enough to meet the requirements of the image processing algorithm, but since the picture will not be seen, then maximizing image quality is somewhat redundant. Instead, the aim is to make the system as light and compact as possible while minimizing power requirements. This SWAP reduction enables more sensors to be deployed in devices with size and weight constraints, such as drones, while also reducing associated costs. As such, small image sensors are typically deployed, often those originally developed for smartphones. Over time, expect to see greater integration, such as stacking the sensing and processing functionalities.   Miniaturized Spectral Sensing Embedded vision isn’t restricted to conventional image sensors that detect RGB pixels in the visible range. Monochromatic sensing will likely suffice for simpler algorithms, such as edge detection for object location, reducing sensor cost. Alternatively, using more sophisticated sensors with additional capabilities arguably supports the essence of embedded vision by minimizing subsequent processing requirements. Adding spectral resolution to image sensors can expedite subsequent processing for applications such as material identification, since the additional spectral dimension generally means a smaller training data set is required. However, meeting the SWAP requirements of most embedded vision systems is challenging for many spectral sensors due to their bulky architectures housing diffractive optics. Emerging approaches aiming to resolve this challenge include MEMS (micro electromechanical systems) spectrometers, increasing optical path length using photonic chips, and adding multiple narrow bandwidth spectral filters to image sensors using conventional semiconductor manufacturing techniques. Event-Based Vision Another sensing approach that aims to minimize subsequent processing requirements is event-based vision. Rather than acquiring images at a constant frame rate, with event-based vision, each pixel reports timestamps that correspond to intensity changes. As such, these sensors combine greater temporal resolution of rapidly changing regions with far less data from static background regions, thus reducing data transfer and subsequent processing requirements. Furthermore, because each pixel acts independently, the dynamic range is increased. Crucially, this alternative acquisition approach is occurring within the sensing chip, not with post-processing algorithms, enabling an associated embedded vision system to be simpler since it has less data to handle. Further Insights The adoption of embedded vision systems will continue to grow as more devices gain autonomous capabilities, with robotics, industrial vision, and vehicles as the dominant markets. This represents an opportunity not just for compact image sensors, optics, and processing ICs, but also for emerging sensor types such as spectral imaging and event-based vision that can reduce processing requirements and thus support embedded vision’s value proposition of reducing size, weight, power, and latency. IDTechEx’s report “Emerging Image Sensor Technologies 2023-2033: Applications and Markets” explores a diverse range of image sensing technologies capable of resolutions and wavelength detection far beyond what is currently attainable. This includes the spectral and event-based sensing outlined here, but also InGaAs alternatives for short-wave infrared (SWIR) imaging, thin film photodetectors, perovskite photodetectors, wavefront imaging, and more. Many of these emerging technologies are expected to gain traction across sectors, including robotics, industrial imaging, healthcare, biometrics, autonomous driving, agriculture, chemical sensing, and food inspection. @IDTechEx #PAuto

Making laboratory miniaturized spectral sensing affordable.

Unsatisfied with smartphone camera color reproduction? Would it be useful for a user to detect different materials or even assess the freshness of food using a phone or other household appliance? The emerging technology of miniaturized spectrometers and multispectral cameras holds the solution. By detecting spectral information rather than just simple RGB color coordinates, these affordable devices can reveal previously available information via specialist equipment.

Acquiring spectral information with affordable sensors is part of a growing trend to detect optical information beyond what the eye can detect. As machine vision capabilities improve, algorithms demand more comprehensive input data than conventional RGB sensors can provide.

IDTechEx’s report, “Emerging Image Sensor Technologies 2023-2033: Applications and Markets”, covers a wide range of these emerging image sensor types, including shortwave infrared (SWIR) sensors based on organic semiconductors, quantum dots or germanium, event-based vision, hyperspectral imaging, quantum imaging and more.

Application Opportunities.
The desire to add more functionality in pursuit of differentiation means that spectral image sensors are in demand for smartphones. In 2021, German research center Fraunhofer ENAS announced its objective to produce infrared spectrometers for smartphone integration that weigh 1 gram and cost just US$1. This is a substantial improvement on the classic infrared spectrometers that weigh several kilograms and cost thousands of dollars. Fraunhofer’s low-cost spectrometer will be able to detect fake medicine and drugs, test air quality, and analyze food decomposition.

A notable player commercializing affordable spectral sensing is Spectricity, an early-stage Belgian company. Earlier this year, it announced its first product – a miniaturized multispectral image sensor that can detect 16 color channels within the visible region. The greater spectral resolution compared to conventional sensors enables more accurate color reproduction, with smartphones the target market given that camera quality is determining factor for many consumer purchase decisions.

Another consumer application that can benefit from spectral sensing is wearable health monitoring. While smartwatches can already track heart rate, oxygen levels, sleep patterns and respiration rates using LED light sensors, but more sophisticated sensors would provide greater functionality. Spectrally resolving wavelengths beyond 1000 nm, for example, by using one of the emerging SWIR (short wave infrared) detection technologies such as quantum dots, enables the detection of biomarkers undetectable by visible and NIR light. Wearable devices could then monitor glucose, lactate, hydration, and alcohol levels, broadening their appeal and providing a source of differentiation amongst the current field of smartwatches with increasingly similar capabilities.

Along with consumer electronics, miniaturized spectrometers and multispectral cameras are expected to gain traction for industrial inspection and robotics since lightweight permits integration into existing equipment. The most promising tasks are object recognition and distinguishing between different materials. Spectral resolution is especially valuable when aiming to distinguish visually similar materials such as different plastics or oil/water.

Outlook.
Due to their compact nature and affordability, miniaturized spectrometers and multispectral sensors promise to bring new capabilities to handheld devices, portable electronics, and industrial imaging inspection equipment. Capabilities that were once the preserve of the laboratory can now be applied to smartwatches, smartphones, and even robotic vacuums.

Miniaturized spectral sensing market in 2033 – by application and detection range

IDTechEx’s report “Emerging Image Sensor Technologies 2023-2033: Applications and Markets” explores a diverse range of image sensing technologies capable of resolutions and wavelength detection far beyond what is currently attainable. Many of these emerging technologies are expected to make waves within numerous sectors, including robotics, industrial imaging, healthcare, biometrics, autonomous driving, agriculture, chemical sensing, and food inspection. IDTechEx expects that the growing importance of autonomous technologies will lead the emerging image sensor market to over US$500 million (€471m) by 2033.

@IDTechEx @AgilityPRS #PAuto #Laboratory #Food

FPGA machine vision.

Innodisk has announced its latest step into the AI market, with the launch of EXMU-X261, an FPGA Machine Vision Platform. Powered by AMD’s Xilinx Kria K26 SOM, which was designed to enable smart city and smart factory applications, Innodisk’s FPGA Machine Vision Platform is set to lead the way for industrial system integrators looking to develop machine vision applications.

Automated defect inspection, a key machine vision application, is an essential technology in modern manufacturing. Automated visual inspection guarantees that the product works as expected and meets specifications. In these cases, it is vital that a fast and highly accurate inspection system is used. Without AI, operators must manually inspect each product, taking an average of three seconds per item. Now, with the help of AI solutions such as Innodisk’s FPGA Machine Vision Platform, product inspection in factories can be automated, and the end result is not only faster and cheaper, but can be completely free of human error.

Innodisk’s FPGA Machine Vision Platform comes with 1GbE LAN, 4 USB 3.1 Gen1 ports, 2 M.2 slots, and a series of other expansion and connectivity options. Thanks to the platform’s 0° to 70 °C operational temperature support, and optional industrial temperature support from -40° to +85°C, EXMU-X261 is tough enough for the harshest of industrial environments. In addition, EXMU-X261 features support for Innodisk’s InnoAgent out-of-band remote management module, allowing the platform to be remotely managed from anywhere, even during a system crash or an in-band network failure. This is important for any automated system, as it allows for it to be completely unmanned, which further reduces manpower and maintenance costs.

Customers can take advantage of AMD’s Xilinx Kria K26 SOM’s ability to quickly get applications to market thanks to EXMU-X261’s full support for Innodisk’s AI Suite SDK. The suite includes an FPGA Model Zoo, as well as Innodisk’s in-house software solutions, such as iCAP (Innodisk Cloud Administration Platform), and iVIT (Innodisk Vision Intelligence Toolkit). iVIT for example, provides a deep learning environment for efficient development and deployment of “no-code-operation” solutions.

@Innodisk_Corp @mepaxIntPR #PAuto #Smart

Spherical image capture.

Teledyne FLIR Integrated Imaging Solutions is pleased to announce the all new Ladybug6— the latest addition to its field proven Ladybug series. Ladybug6 is the leading high-resolution camera designed to capture 360-degree spherical images from moving platforms in all-weather conditions. Its industrial grade design and out-of-the-box factory calibration produces 72 Megapixel (MP) images with pixel values that are spatially accurate within +/- 2 mm at 10-meter distance.

Key Features: • Captures 72 Megapixel images with spatial accuracy of +/- 2 mm at 10-meter distance • Frame rates of up to 29.9 FPS at 4K resolution, 15 FPS at 72 MP • Wide operating temperature range of -30° C to 50° C • IP65 rating with industrial grade IP67 rated connectors • Support for additional Global Navigation Satellite Systems • Excellent color response, low noise, and a high dynamic range • Captures, compresses, and transmits 8-bit or 12-bit pixel data • Feature rich and user-friendly Ladybug SDK • All-metal body with 2-year warranty.

“The new Teledyne Ladybug6 is designed for mobile mapping and all-weather inspection projects requiring excellent image quality and high resolution,” said Mike Lee, Senior Product Manager at Teledyne FLIR. “With the addition of Ladybug6, we are now pleased to offer a wider variety of spherical cameras with higher resolutions ranging from 30 MP to 72 MP.”

The new Ladybug6 builds on Teledyne’s machine vision heritage with increased image resolution, enhanced on-board processing, and robust IP67-rated connectors. Building on the field proven Ladybug5+, the Ladybug6 captures, compresses, and transmits 8-bit or 12-bit pixel data delivering outstanding images across a wide range of lighting conditions with excellent color response, low noise, and a high dynamic range. Designed from the ground up to capture images from moving platforms in outdoor environments, the Ladybug6 features a wide operating temperature range (-30° C to 50° C), support for additional Global Navigation Satellite Systems, and trigger control by hardware or software with advanced APIs for complete camera control.

Ladybug6 cameras are engineered to deliver high-accuracy, high-resolution, and dependable results for applications such as HD mapping, asset management, roadside inspection, panoramic street image production for street view, road surveying, heritage scanning, building management, among several others.

@flir @mepaxIntPR #PAuto #Inspection

How to quickly prototype 4x machine vision applications on one small embedded system.

Embedded vision components are ever popular and being incorporated into a plethora of applications. What all these applications have in common is the need to pack more functionality into tight spaces.

Often, it is advantageous for these systems make decisions on the edge. To enable such systems, including the ability to prototype quickly, Teledyne FLIR has introduced the Quartet™ Embedded Solution for TX2. 

This customized carrier board enables easy integration of up to 4 x USB3 machine vision cameras at full bandwidth.

To highlight what the Quartet with Spinnaker SDK pre-installed can enable, we describe steps taken in developing an ITS (traffic systems) inspired prototype running four simultaneous applications, three of which use deep learning for:
• Application 1: License plate recognition using deep learning
• Application 2: Vehicle type categorization using deep learning
• Application 3: Vehicle color classification using deep learning
• Application 4: See through windshield (past reflection & glare)

This very practical paper includes (1) a shopping list, (2) development time for each application, (3) number of training images required, and finally, actions to optimize overall system performance.

@flir @mepaxIntPR #PAuto

Light industrial camera.

The latest additions to the FLIR Blackfly S GigE camera line – the BFS-PGE-50S4M-C (Mono) and BFS-PGE-50S4C-C  (Colour) have been introduced. These 5 MP models are particularly well suited for integration into small handheld devices with their impressive low weight of 53 grams and high pixel density ideal for integrating with compact, less expense lenses. Leveraging Sony’s IMX547 sensor they deliver exceptionally low-light performance with superior quantum efficiency and very low absolute sensitivity making them suitable for a range of challenging applications from biometrics to scientific research and more.

These new GigE cameras with power over ethernet also take advantage of their Lossless Compression feature, delivering 25% higher frame rate of 30 FPS without compromising on image quality. Read along for more detailed specs, pricing and to consult a specialist about your project requirements.

• 5 MP global shutter CMOS
• At 53 grams, the lightest weight camera with IMX547
• Lossless Compression enables 25% higher FPS (from 24FPS to 30 FPS @ full resolution)
• High QE and low noise yield a low Absolute Sensitivity Threshold (4.4 photon AST)
• Superior quantum efficiency (68% QE @ 525nm)
• Excellent low-light performance (2.49 e- read noise)
• Small pixel size enables less expensive, more compact optics
• High sensitivity (High QE and Low AST) reduces lighting system requirements

Customization Options and Special Requests
Researchers, manufacturers, and diagnosticians in specialized fields know the smaller the scale, the more important it is imaging devices are free from any particles that might obstruct or blur a magnified sample. Teledyne FLIR cleans and assembles all its machine vision camera optical assemblies in an ISO certified clean environment (ISO7 - Class 10000). This standard level of dust control is sufficient for most uses however, some applications require an even higher standard. In such cases, their Enhanced Dust Control Service is available on all area scan machine vision cameras except Blackfly S Board Level (USB3/GigE) and Firefly models.

@flir @mepaxIntPR #PAuto 

Ensuring top quality cotton swabs!

Smart Manufacturing Integrates Machine Vision and Industrial Robots 

 

Many people cannot imagine a day without using cotton swabs. Behind such a small product, there is extensive know-how from production to packaging. Since the unit price is relatively low, the investment in time and resources is highly sensitive. To optimise results, the cotton swabs must be quickly placed into a package in order. This is a challenge for traditional manufacturers that mainly rely on manual labor.

Yet another challenge, is that there is an ever increasing shortage of manpower in the manufacturing industry. To enhance productivity and maintain a consistent product quality, many manufacturers adopt automated equipment and processes. On top of that, smart manufacturing is becoming an unstoppable trend. To prepare for this new era, building highly automated production lines has become a necessary move for the future.

Delta recently built a cotton swabs’ automatic inspection and packaging solution for a packaging partner working with a cotton swab manufacturer. The solution integrates a vibration bowl, conveyor belts, industrial robot, and machine vision for an all-in-one automatic machine. It improves overall performance with higher productivity and longer runtime. In addition, the adoption of machine vision avoids the inconsistencies of manual inspection and ensures product quality.

The solution includes the High-Performance Slim PLC DVP-SV2 Series that delivers cotton swabs from a vibration bowl and transfers them to the inspection spot under the Machine Vision System DMV2000 Series. The SCARA Robot DRS40L Series then tracks the conveyor and picks-and-places the qualified ones into the case in a set order. While the robot sends a signal to the PLC for counting, the pick-and-place movement stops when it reaches the set number. The PLC controls the positioning cylinder to switch to a new case and enable a smooth process for packaging and inspection.

Delta cotton swabs automatic inspection and packaging solution provides the following benefits and advantages for the manufacturer and users:

• Automated production line enhances productivity The automated production line integrates the 10.1-inch Widescreen Ethernet HMI DOP-110 Series, High-Performance Slim PLC DVP-SV2 Series, and SCARA Robot DRS40L Series. It extends runtime and enhances productivity. • Machine vision maintains consistent product quality Unlike manual inspection, the Machine Vision System DMV2000 Series avoids distraction and can quickly and precisely detect defects and maintain the consistency of product quality. • Quick pick-and-place ensures the package counts Delta’s SCARA Robot DRS40L Series quickly tracks the conveyor and sends signals to the High-performance Slim PLC DVP-SV2 Series to count whether the quantity of cotton swabs has reached the set number, and ensures the accuracy of each case.

Delta’s cotton swab automatic inspection and packaging solution integrates PLC, machine vision, and industrial robot in an all-in-one machine. It decreases the need for space and is capable of quick and precise packaging. Besides, it helps users overcome labor shortage challenges. The solution is highly automated and flexible for industrial 4.0, which may help users embrace the potential and opportunities of smart manufacturing.

@DeltaEMEA @mepaxIntPR #PAuto #Robotics

Machine vision board changes.

The Supervisory Board of Stemmer Imaging has appointed Uwe Kemm to the company’s Executive Board effective from 01 April 2020. As a member of the Executive Board (COO) he will be responsible for the further development and implementation of key initiatives in the context of corporate strategy in addition to overseeing operations in key business units.

Uwe Kemm

Uwe Kemm has more than 30 years of experience as a manager in international and national companies in the technology industry. His responsibilities included, in particular sales and business development as well as successful market launches of new products and technologies. In addition he has assumed responsibility for various company acquisitions as well as organizational development, integration and change management.

At the same time, Martin Kersting (CTO) resigned from the Executive Board as of 31 March 2020 for personal reasons and is leaving the company at his own request. However, he will remain involved in an active advisory function.

"On behalf of the Supervisory Board, I would like to thank Martin Kersting for his many years of very dedicated service to the company. In Uwe Kemm, we have gained a technology expert with many years of management experience with both a global player and with a number of medium-sized companies in the software market," said Klaus Weinmann, Chairman of the Supervisory Board of Stemmer Imaging AG. "We wish him continued success and look forward to working with him in a trustful cooperation".

"Since summer 2019, Uwe Kemm has provided key inputs and achieved decisive results in strategic projects in a consultancy capacity for the company. I am looking forward to a successful cooperation",  added Arne Dehn (CEO)

#PAuto @stemmerimaging

Machine vision programming upgrade!

The release of CVB 2019 introduces new APIs to provide an innovative way of developing machine vision solutions with Common Vision Blox, together with an impressive range of new tools. These tools provide functionality for hyperspectral and polarisation imaging, as well as OPC UA machine to machine communication capability for Industry 4.0 requirements. In addition it is now possible to develop machine learning solutions on embedded platforms. Object recognition speeds have been greatly increased, as have video sequence recording speeds.

CVB Image Manager and CameraSuite 2019 feature three new object-oriented APIs, supporting C++, .NET and Python. These APIs make getting started with Common Vision Blox significantly easier and help eliminate some of the most common programming issues in lifetime and memory management. Python support in particular brings the power and speed of CVB to a new group of users who are used to quick results and easy programming. The new APIs are compatible with the existing CVB API, meaning that users can build on existing applications without having to re-write their codebase. Linux support now covers Ubuntu 18.04 as well as 16.04 on Intel platforms and ARM. CVB software protection (WIBU dongles and CameraSuite licence) is also now working on ARM.

The CVB Foundation Package 2019 now benefits from a communication framework for Industry 4.0 (OPC UA) and Polarisation functionality as well as improvements to the speed and stability of 3D calibration in Metric. The OPC UA framework allows OPC UA server and client functionality. It is useful for machine-to-machine communications between vision systems and robots or PLCs, for example and is an integral part of an Industry 4.0 connected factory. Polarization is a full toolbox to get the most from the polarisation cameras. It allows easy access to the data from the polarisation sensors from Teledyne DALSA and Sony (used in many camera families). Everything from simple ‘de-mosaicing’ to ‘minimum reflection images’ to Stokes’ parameters and the angle/degree of polarisation can be handled, enabling reflectivity and surface-normal calculations to be made for applications in a wide range of industries.

CVB Spectral is a completely new tool to help industrial users harness the power of hyperspectral imaging for scientific, wavelength-based analysis. CVB Spectral is a complete toolbox for acquiring and preparing hyperspectral data from hyperspectral cameras and ENVI files so that traditional image processing tools can be used on the results. A hyperspectral data-cube, which can also be acquired, saved and loaded using CVB Spectral, can be corrected for optical and lighting effects to create wavelength-correct RGB or LAB colour data. This gives real area colorimetry information far more accurately than can be achieved with 3-chip cameras. The data-cube can also be sliced to produce useful 2D images for further processing. Typically these might be (x, wavelength) or (y,wavelength) at a certain x or y value.

CVB 2019 also contains some significant updates to existing tools. Recognising the importance of embedded vision and machine learning as key machine vision themes, the CVB Polimago machine learning tool is now available for Linux on Intel and ARM platforms. A raw recording method has been added to CVB Movie. This makes full use of the speed of the disks and the data-formats of the camera to allow recording speeds of gigabytes per second. In addition, CUDA support has been added to the 64-bit Windows version of CVB ShapeFinder. This means that with the right GPU hardware it is possible to increase search speed by 250%. An NVidia Jetson version is also available on request.

#PAuto @stemmerimaging