Adversary Based Design

Basics

• A Method of Review wherein the Product/Policy Proposal is approached via a Reviewing Party acting as the "Adversary"
• The Adversary's goal is to break the device/system
• For instance if the product being tested was a paper shredder, they would figure out how to make it fail, as well as ways to injure oneself/cause damage (within reason) using the device
• If it were a Video Game, they would do everything they could do to "Break the Game"
• If it were a Tax code, they would figure out every loophole they could use to their advantage
• An Currently Existing Example would be Penetration Testing for (Cyber-) Security Applications
  • This entails a group hiring companies/individuals who specialize in "pen-testing" to essentially hack their system, so they can then repair any exposed bugs / vulnerabilities
    • A Related concept is "White Hat Hacking"
  • It can also be done for Physical Security (Ie sneak into / break into this building)
• Another good example is Lockpicking / Safecracking (See Lock Picking Lawyer (YouTube Channel ), which, at least in the case of safes, is done by the company to determining how long it would take to destructively "cut into" a safe etc
• It Does Differ to Existing Implementations in so far that the adversary Based Testing / Destructive Testing plays a far more active role in the design process
• ie rather that "We are finished with the design, off to testing"; from day one the design team is thinking "How can i break this / how could someone break it" thus it can also be more of a Design Philosophy at times rather than purely a testing protocol

Well Documented Examples

Stuff Made Here (YouTube Channel) + LockPickingLawyer (YouTube Channel) 's "Unpickable Lock" Project Collaboration

• "Stuff Made Here", a channel known for various goofy, yet intricately designed, contraptions released This Video in which they designed a lock that should not be "pickable" (At least via conventional means)
• They then got a commercial locksmith to attempt to pick it, which they were unsuccessful in doing (albeit in real world conditions; it was in a door, and no knowledge beforehand of it's inner workings/special self build tools were used)
• They then sent off a version of the device to the YouTube Channel "LockPickingLawyer" (With This Video detailing a second improved version made pre-testing as well that mainly improved reliability of the mechanism / reduced jamming)
• It was tested by the LockPickingLawyer in This Video

MihaiDesigns's Constructive Criticism of the PrusaXL

• A Video by the YouTube Channel "MihaiDesigns" Titled "Original Prusa XL in depth impressions" (~22 Minute Watch)
• If they created Mods for it with their ideas, or simply a fork / different printer, this would be a great example, especially since Prusa Research tends to cross communicate community wise

Past OSE Examples

• This may fit better as some sort of "Real World Use Test" / "Demo Usage" concept/page, as it was less intentional / *adversarial*, but still fulfills the same goal to an extent

CEB Press (Need to grab what version was used) and the Belize Build

• (Need to grab the video link, but some of the valve switching failed (if i remember correctly)

Internal Links

• Competition Based Design
• Rapid Iteration / RITE Method
• Destructive Testing
• Evidence Based Charity / Evidence Based Philanthropy
• Peer Review
• Review

External Links

• The Wikipedia Page on Adversary Evaluation (A Related Concept, More Related to Policy Decisions)