BLOG
BLOG
2022/10/07 | Time to read: 3 min
Dalia Peña is the head of customer experience at Drishti. Dalia brings extensive experience and a keen understanding of lean manufacturing principles to Drishti. She has been able to infuse that background into the company, particularly to members of Drishti who bring less manufacturing expertise to the table. In doing so, Dalia has ensured that Drishti’s AI-powered computer vision technology stays hyper-focused on meeting the needs of discrete manufacturers across North America. Dalia was one of Drishti’s first employees, and at the time, one of the only employees with actual assembly line experience.
In manufacturing, we attempt to control as many variables as possible in order to limit the number of places where things can go wrong on the production line. We create processes that are prescribed and repeatable and produce consistent results in both quality, quantity and time. Mature manufacturers have this process down to a science. But how do we know that a process is actually good? How do we know that a change we’ve implemented will not have adverse effects on production? The answer, of course, is verification.
First, a few simple definitions to make sure we’re all on the same page.
Verification is ensuring that a product is made within the specifications and guidelines of the product requirements, as well as any policies, regulatory or otherwise that may be needed.
Validation is actually ensuring that the end product itself is acceptable. It is, in effect, measuring the end result. For the purposes of this discussion, we’re speaking mostly about verification. Verification is a key part of manufacturing and a vital companion to standardized work. As standardized work changes and evolves, verification does the important job of confirming that the process is still functioning as desired.
Assuming that the product consistently validates within specification, the best way to ensure continued success despite constant change is to bolster verification procedures. If a verification process is robust, it becomes much easier to track down issues when they occur.
Verification is built into the change procedures of mature manufacturers. The process of implementing new or changed processes on the line is captured in the plan do check act (PDCA) cycle. The “check” part of the process should not be understated. In the check phase, once again, we verify and validate that the process is performing as expected and that the resulting product is appropriate.
Here’s a common scenario: Especially when a new process is developed, verification is performed successfully in the moment under the watchful eye of trainers, quality staff or industrial engineers, but then the process inevitably veers over time, and verification doesn’t follow suit. The solution would be to enact continuous verification of the process, which would ensure that the methods prescribed for a product to be built were being carried out correctly. In industries with high risk involved, like medical devices or aerospace, an end product failure could mean catastrophe (including death or serious injury).
To be fair, manufacturers do re-verify processes per their policies periodically in order to stay certified by regulatory bodies (such as ISO and the FDA). But this is still more of an act of compliance than qualitative assurance. In order to truly reduce risk and ensure quality of process, all manufacturers would need to verify far more frequently than is common practice.
The reason they don’t is because of the resource heavy and burdensome requirements continuous verification would require — making it unrealistic for an organization to commit to. Even certifying bodies limit their requirements for verification and validation because of the burden it presents to the line. It is unrealistic to assume qualified personnel could provide continuous verification of process for all of production indefinitely. The net result is that when product quality issues do occur, they are usually caught after the fact and subject matter experts have to trace breadcrumbs back to try and identify the source of the problem.
Luckily, new technologies can bring capabilities to manufacturing that never before existed, making continuous process verification possible for the first time. Consider Drishti — Drishti helps to alleviate risk by continuously verifying that processes are being followed as prescribed. Drishti’s patented action recognition can provide insights in real time whenever there is a deviation from the verified process.
With Drishti, the production line is experiencing constant verification, and thus if a product has a problem, we can trace it back to its build and quickly see whether the process is to blame. Drishti supplements manufacturing experts by providing data, allowing operations, engineers and quality experts to focus on problem solving rather than problem finding.
What this level of verification means is peace of mind. High risk products can be continuously verified and alleviate any concerns for recalls or manufacturer wrongdoing with video evidence available on Drishti’s platform. Subject matter experts will be thankful for the support and extra time saved when Drishti is at work.
