Rockville MD, January 17, 2020 –(PR.com).– Acquired data Solutions (ADS), a leading provider of Computer Vision (CV), is pleased to introduce Transform-CV. This system, which is a COTS and deep learning enabled Computer Vision (CV), will be used for visual quality inspection. Transform-CV, designed in partnership with RADX Technologies provides a modular platform that is low-cost, flexible and easy to tailor. Manufacturers, distributors and integrators from all industries and sizes can use it to improve their component and subsystem Quality Assurance accuracy, throughput, and reliability.
The majority of small and medium-sized manufacturers rely on manual Q/A inspection processes. These processes are usually dependent on personnel and, therefore, have limited repeatability, reproducibility and scalability. SMB manufacturers often find logic-based CVs too expensive, especially when it comes to high-mix and low-volume applications. Traditional CV systems are also usually proprietary and inflexible. They require initial programming by CV specialists and reprogramming when new “devices to be tested” (DUTs) or features or defects are added. This time-consuming and expensive process of initial deployment and updating is often a tedious and expensive one.
Transform-CV, on the other hand, is based upon COTS components, and uses deep-learning, leveraging NVIDIA’s GPUs to accelerate both the training process and inference operations. This results in a cost-optimized high-performance solution which can be “trained”using images or drawings of DUTs from customers and their relevant features or defect (provided by Q/A or assembly personnel). Transform-CV, once trained, can be run in real-time by an intuitive GUI. This greatly improves the repeatability and reproducibility of Q/A operations. Transform-CV also supports customer-enabled learning. By collecting and annotating images, users can โteachโ Transform-CV how to operate on new DUTs, or recognize new features or defect without requiring any additional programming or consultation by ADS or other third parties. This dramatically reduces the Total Cost of Ownership. Transform-CV can also be used on existing PXIe and PCIe systems by customers, reducing acquisition costs.
Written in Python and PyTorch, Transform-CV can be used on Linux or Windows-based PXle Systems and PCs. It can also easily interface with MATLAB LabVIEW and other popular application frameworks. Transform-CV is capable of a wide range of inspections, including macro- and microscopic dimensional analyses, workmanship inspections, sorting and classifying, and defect detection. ADS is developing Transform CV systems for microelectronics applications, cable manufacturing, kitting CNC machining and additive manufacturing.
The typical Transform CV system is composed of the following components:
* The Transform CV Base License includes a customizable framework for training, inference, and GUI.
A PXle- or PCle based embedded computer vision systems equipped with a suitable NVIDIA GPU, and other essential components. Or a GPU kit for existing PXle- or PCle based systems.
One or more camera(s) for application-specific use with appropriate optical training, including magnification, filters and illuminators.
A manual, semi-manual, or automated fixture to hold the DUTs if robotic or hand-held operation is not used.
* Application-specific tailoring, including system training using images and/or diagrams and incorporating pass/fail criteria for the required DUTs.
Optional integration with CRM or ERP systems.
Installation, delivery, and personnel training regarding how to add DUTs, Features, etc.
Pricing, Availability and Availability. Transform-CV Basic Inference and Training COTS Binary Right to Use License starts at $15k with optional annual updates. Transform-CV systems are priced from $25k to up to $500,000, depending on the application, required hardware, software, and tailoring. Delivery time varies from 3 to 12 month depending on the complexity of the application.
Steve Seiden is the President of ADS. “ADS worked closely with RADX, and key launch partners, to develop Transform CV, which we believe will bring this powerful yet cost-effective tech to small and mid-sized businesses looking to improve their Q/A operations or scale them up,” said Steve Seiden. “CV systems are expensive and rigid. This has limited their use in high-volume, low-mix applications. Transform CV, a COTS AI-based system that is inexpensive and can be used in low-mix, high-volume applications.
Contact ADS to learn more about Transform CV’s applications and to see how it can improve Q/A performance.
Acquired data Solutions (ADS) has over 25 years of experience in delivering integrated solutions to federal and commercial clients for test, process automation and cybersecurity. Our solutions are designed with the highest security standards, and always exceed client requirements. We offer a range of services, including System Integrations, Legacy Upgrades and Software Developments, Cybersecurity Risk Assessments and Program Management Support. Our mission as an ISO 9001-certified organization is to provide high quality products and services that align with regulatory and customer requirements.
RADX Technologies, Inc., a high-tech, small business, develops advanced, COTS, HPC technologies and products for PXle and PCle based Modular Test and Measurement applications (T&M), Electronic Warfare, Automotive ADSAS/AD, Manufacturing Quality Assurance and Semiconductor Applications. RADX has its headquarters in Palo Alto, CA, with development and production locations in California and Taiwan.
Acquired Data Solutions, Inc.
Jean Bartlett
301.984.4193
www.acquiredata.com
Contact
Categories
- Business
- Computer Hardware (
- Computer Programming]
- ]Engineering
- Manufacturing Technology