Instrumental Spills the Beans on Taco Assembly

Instrumental Spills the Beans on Taco Assembly

Isaac Sukin
April 01, 2017

Here at Instrumental, a core cultural value is our love of food. That’s why we've recently expanded our operations into the taco vertical. In this post, we'll be spilling the beans on how Instrumental has improved the taco assembly line.

This dataset of work-in-progress taco photos comes from a recent in-house build set up to dog-food our own technology. After EVT and DVT we were rapidly approaching product lunch. We augmented our assembly line with Instrumental stations to take photos at six high-risk steps:

  1. Enclosure
  2. Rice & Beans
  3. Protein
  4. Toppings
  5. Sauces
  6. Finished Good
Taco Assembly

Our line leader Sam instructed us on the strict SOP and provided us with serialized barcodes to scan for each unit. Then we went to work!

Taco Barcodes

When our product design engineers reviewed the images in Instrumental’s web app, it quickly became clear that we had some anomalies. For example, this Finished Good photo shows an extra part! We knew we would bean trouble if this was a trend.

Fork Anomaly

At first glance, we suspected an upstream vendor problem, but this lettuce down the wrong path. By zooming in on the high resolution, 24 megapixel image, we determined that the fork was misplaced by an operator, and so we implemented corrective action in the form of improved operator training and a simplified operator forkflow.

Another time-consuming issue that filled up our images was that some tacos had a bad queso missing meat:

Veggie

After remotely inspecting this unit, we determined that this condition is acceptable, and adjusted our spec accordingly. Being able to confirm via images that this taco was not problematic saved us valuable taco-eating time.

Because of the issues we found, we decided to run a DOE—another round of tacos to make sure we had no unresolved issues.

Doe

Thanks to our technology and processes, we're happy to report that our lunch was burri-totally successful!

Manufacturing Intelligence

Don’t miss a post, subscribe today.