Most advanced manufacturers of appliances have their own set of production tools that they use as standard – including standard work in their factories that outline work steps, expectations, etc. They try to build quality into the process (through things like poke-yoke) and have some pretty good systems for this, but there are places where this is hard. Specifically, adjusting mistake-proofing and quality inspection systems to different configurations coming down the same line – as many of their products/lines have to customize to customer specific orders.
A key challenge for automated quality control in mass-customized production is the need for 1000s of “training images” for every product variation coming down the line. We eliminate this need by defining compositional product models where machine learning algorithms are trained on sub-assemblies and combined together for individual build orders extracted from the Manufacturing Execution System (MES).
We work with manufacturers to integrate our solution into existing MES systems, or provide a custom barcode solution where our platform identifies each individual build order.
Our Pathfinder platform builds compositional models of a variable physical assembly task, given only a handful of recorded demonstrations of the task. Once such a model is built, Pathfinder can finely track the job status from live video, to guide a worker through the task, provide independent training, and perform analytics. Instead of requiring a team of specialized computer vision engineers to build custom interactive experiences that can track the steps a worker successfully performed which takes months, we enable such AI-driven feedback to be set up in hours by ordinary manufacturing engineers.
Customers use Pathfinder to verify the assembly of compact subassemblies which are built in layers, e.g. planetary gears, valves, gas filters, and different kinds of motors and thus traditional machine vision solutions are irrelevant for product defect avoidance. Pathfinder is deployed for both standards compliance as well as a digital Poka Yoke tool, and routinely results in 85% or more reduction in manual assembly issues. Customers reports product returns down to 20% (due to missing washers, keys, and such) of their previous levels upon Pathfinder deployment.
Pathfinder has also been deployed in Kit Packing applications in the hardware industry where it is important to include multiple components and manuals in a single package, e.g. when packaging a Drilling machine. Pathfinder deployments have resulted in complaints minimize to under 10% due to missing manuals or chargers in these situations.
Due to very high very high turn-over rate, manufacturers are having to train tens to hundreds of new assembly workers every month. Training costs are enormous, and the lack of training opportunities leads to defects on the production line. Pathfinder is used both as a direct training tool as well as to measure the proficiency of new operators to speed up operator onboarding. We have consistently demonstrated that we can halve the time needed to onboard workers in highly sophisticated scenarios. In fact, our projected guidance tool is built to enable operators to be productive within minutes of arriving at a workstation for the first time.
Huy takes our product roadmap and figures out what’s achievable with today’s AI advances, and what we need to invent to solve our customers’ pain points.
Huy led teams of scientists and engineers at NEC and invented technologies for driver assistance systems that are part of 500,000+ Honda cars. He is an inventor on more than 20 US patents, and his research papers are cited by every lab from MIT, Stanford, and CMU to Apple, Google, and Baidu.
Michael works closely with our marquis customers to understand their needs, and figures out the best solution that immediately adds dollars to their bottom line.
Michael has almost 20 years of enterprise sales experience. He has held sales leadership roles at pioneering AI and computer vision startups.
Andrey takes inventions and converts them into innovations. He leads our product engineering team to turn our scientific prototypes into robust customer-facing products.
Andrey co-founded and successfully exited a startup to immigrate to the US. He spent his next eight years building the core data science infrastructure for Amazon’s supply chain planning team, and the CV/ML infrastructure for Microsoft’s HoloLens organization.
Zeeshan focuses on delighting our customers by understanding their needs and defining product and strategy to meet and exceed these needs. He works closely with our sales leadership to grow our customer base.
Zeeshan was a scientific leader at Microsoft’s HoloLens division, where he invented AI technologies to assist frontline workers across industries. Earlier, he shipped IoT platforms and visual defect detection software for manufacturers at Siemens and MVTec GmbH.
