Secure DevOps for PLM

O brother, where is thy Stinger?

I always support the right honorable Bryan Finster in his outbursts against government-associated lack of efficiency compared to the business world. Though compared to Bryan’s experience in the field, I cannot fully comprehend how bad things are and what can be done to improve them, but these days I seem to observe a fairly good example.

As you may have heard, there is a war going on in Ukraine. This war will not end all wars, but it will definitely redefine them for the foreseeable future. In many ways, it is redefining 21st century military conflict back to WWI and WWII levels of intensity and consumption of war materiel.

Neither Iraq, Afghanistan and Syria, nor any of the latest Israeli confrontations are equivalent to what is happening in Ukraine now. The closest comparison is the Yom Kippur war back in October 1973. The situation with equipment and munitions became so desperate then that the US had to organize a continuous air bridge to resupply Israeli forces – which were being devastated by Egyptians wielding their unexpectedly effective Soviet-made weapons. Presently the Ukrainians are putting up stiff resistance while being totally outgunned in terms of tanks, artillery, aircraft and ubiquitous UAVs. A big part of the credit goes to Javelins countering Russian tanks, while Russian aircraft and UAVs are frequently and successfully attended by Stingers.

The problem with Stingers is that there are not enough of them. Based on the numbers being discussed, the US has already shared with Ukraine significant portion of its own stock and now needs to replenish the inventory urgently.

However, this is easier said than done. The original Stinger components will officially “…become obsolete in 2023, and life extension will keep them until 2030.” Moreover, some components “…are not being manufactured at all and …will need to be redesigned.”

As a matter of curiosity, I decided to check publicly available information about the processes Raytheon was using for Stingers. According to Google:

Document designs using CREO and CATIA CAD software packages to generate engineering drawings. Work with SMA Program Management, Configuration Management, and Program Reviewers to release and archive drawings per Program, Customer, and Raytheon requirements.

In other words, Stinger design and manufacturing are still being maintained with a previous generation of CAD design methods and tools. This compares rather unfavorably to large commercial aerospace OEMs whose 3D model based design (3D MBD) data is a principal element of their strategic productivity drive, including streamlining their type certification process. Productivity improvements come along with the efficiencies throughout the production process which ultimately affects their bottom line.

The government is not in that same spot. They do not have a business mindset, nor do they have to worry about being profitable. Companies like Raytheon understand this well, and they can always spool up a cost-plus army of engineers for a massive new contract or an expansion of an existing one.

If Raytheon needs to redesign certain components, how will they ensure that these components satisfy the same form-fit-function? Google says Raytheon’s Stinger program uses document-based IBM DOORS for their requirements management, which means they may get certain headaches reconciling that 20+ years old stuff with the modern technology landscape. Considering the size of the funding package, Raytheon will certainly find its way out of this predicament, but you cannot always fight fire with more money. Failure to maintain a comprehensive SysML based requirements model has a potential to turn any Stinger-like situation into a costly and not necessarily happy-ended adventure.

Finally, whatever new system will be developed to replace Stinger, it will certainly be much more software intensive than its predecessor. Therefore its development must utilize Agile and DevSecOps best practices to move “at the pace of relevance” – otherwise survival of the Western species may not be guaranteed. We at Senticore are voting strongly with all our fingers, toes and brains combined for such a survival and will be happy to support it. Naturally, side by side with Bryan Finster, Nicolas Chaillan and the rest of the right honorable crowd.

P.S. A few useful links:

  • How to deal with legacy design data in the aerospace domain: Gulfstream Aerospace digitization project.
  • A great explanation by Jos Voskuil why 3D MBD is a good thing.
  • An historically interesting article about CAD data interoperability circa 1997.
  • A university project (sic!) to create anti-UAV portable weapon.