The data that takes the luck out of EVT, DVT, & PVT

Leslie Williams

Instrumental’s mission is to enable manufacturing teams to build better. If you trace back to where the pain starts in manufacturing, it’s in development, commonly called the New Product Introduction (NPI) phase. The NPI process that the most-admired companies in the world use is fundamentally flawed: it relies too much on luck and not enough on data. As a result it’s slow and causes major inefficiencies that cost companies hundreds of thousands or even millions of dollars. Instrumental customers are able to build better in NPI by ditching luck for reliable issue discovery, and providing access to organized, traceable, and real-time data.

During the EVT, DVT, and PVT builds, engineers run a repeating loop of issue discovery, root cause analysis, corrective action, and validation. It starts with engineers toiling to discover issues, often requiring flights to factories and physically walking the line with the hope that they can be in the right place at the right time to catch something amiss. Once an issue is discovered, the real work begins. Engineers work through the challenges of finding root cause with minimal data – they rely on factory communication, human tallied reports, and remote-controlled teardowns (if they aren’t in the factory). Even after they have implemented a corrective action, closing the loop and validating that the fix truly worked is tricky, because, yet again: minimal data. Many engineers lack real-time factory visibility and holistic reporting, so validation is sometimes an afterthought in favor of the next burning discovery.

Instrumental has a better way: automatic issue discovery, immediate access to the data you’d need to root cause a problem, and a way to know that the issue won’t come back. Instrumental optimizes NPI by accelerating each stage of the failure analysis loop.

Accelerate issue discovery with AI

Issue discovery is arguably the most important part of the failure analysis loop because you can’t fix an issue until you are aware of it. Instrumental accelerates issue discovery with AI that is purposely built for manufacturing environments. With just 30 units, and no “perfect” or “bad” examples, users can set up their own algorithms to discover defects and monitor for issues. Instrumental AI is also unique in that it catches unanticipated defects and is not limited to particular failure modes, which is accomplished by intelligently sorting anomalous units within the existing dataset and allowing users to decide what level of variation is defective. This means that on day one of the build, engineers can use Instrumental AI to begin discovering defects and building their issues list without being in the factory – cutting significant time out of the FA loop.

Engineers leverage AI to discover issues in a matter of clicks.

Narrow down your search for root cause

Whether you are investigating a known issue or using Instrumental AI to surface unanticipated defects, you have powerful information to narrow down your search for root cause. Multiple Instrumental stations at key states of assembly allow you to pinpoint where in the line the defect originated, as opposed to just relying on end-of-the-line functional test data which provides limited direction. Instrumental’s image data also enables virtual teardowns, making it easier to figure out if an issue source is design, process, or part related – and to prove it with images.

Rylo Tear Down Stages
Conduct virtual teardowns with images of every unit at key states of assembly.

Communicate with data, spur action

High quality data like high-resolution images and statistics about defect rates provide the best tools for productive conversations with factories, suppliers, teammates, and management. Data keeps conversations factual rather than speculative and images make communicating defects across language barriers easier. Instrumental provides images and data in real-time, so you can quickly gain context, take action, and communicate efficiently without waiting on human compiled spreadsheets from factories.

pareto image with screws
Present top issues contextually.

Validate corrective action with AI

The thorough validation of corrective action is reliant on continuous reporting. For instance, if the corrective action involved an SOP update for an operator action, you would want to ensure that the defect doesn’t resurface after the initial training and any operator turnover – ideally monitoring for this defect on an ongoing basis and flagging it if it resurfaces. Instrumental uses AI to continually monitor for and flag defective units. If you find and fix something in EVT, but the issue comes back in DVT, you’ll be notified immediately.

intercept trends
Continually monitor defects and drill down when needed.

Fast cycles, smooth ramps, mature products

Instrumental clears protracting bottlenecks and accelerates the entire failure analysis loop. Engineers find and fix more issues, have more data to work with, and have the ability to work remotely. The repercussions during NPI are faster cycles, smoother ramps, and reduced costs – ultimately delivering ROI that is hard for leaders to ignore.

Related Topics