Note : This is a fairly technical post, check here if you’re not an IT professional.
Both manual and computer-assisted asset management systems typically require the assignment of a unique identification code to each asset.
This ID has two main responsibilities :
- Be unique: make it easy to refer to each asset in a standardized, unambiguous way
- Be understandable: encode useful information about the asset (owning department, type, acquisition year, etc.)
It is no so easy to be unique
Being unique can be challenging when new codes are manually assigned. It may require a single person with the ability to assign new IDs with the same prefix, and distinguished by sequence numbers. It may start generating duplicates when unicity assumptions (say, acquisition time) are no longer true because of scale. It remains vulnerable to human mistakes. Etc.
However, computer-issued new codes can have unicity issues too. They usually make it impossible to generate new IDs during IT outages. When they outgrow a single machine, they become tricky to get right in online systems, because distributed consensus is never easy.
Being understable is easy; remaining so is unlikely
Every attribute of a fixed asset, encoded in its ID, that can change, will change. This is an iron law. Even attributes that, in principle, cannot change are still subject to human error during input, and changes in format as the business grows.
In the best cases, an attribute change results in changing the ID, and updating it in all places it is currently used. This is an uphill battle, with plenty of opportunities to make the situation worse. It is rarely done.
Most organizations keep the initial ID, with various unstructured notes saying, in essence, “we know this ID means this chair is used by the accounting department, but trust us, it is used by engineering now”. This is even worse : the code is now actively misleading, requiring counteless mental gymnastics, which defeats the original goal of being understandable. This meaning-drift happens in most organizations, and is considered a fact of life.
Why do we keep falling into this predictable mess?
The root cause of unicity issues is that having multiple ID issuers that need to coordinate to guarantee unicity is simply too onerous for most realistic business scenarios. The only way out is to eliminate this requirement.
Meaning-drift happens because the real world changes all the time, and trying to capture it is doomed to fail. Finding another way to make asset attributes easy to access, for a given ID, will remove the temptation of putting the attributes in the ID itself.
QR codes and global IDs to the rescue
Global IDs like UUID can be independently generated, offline, by different departments, and are mathematically guaranteed to be unique. The solve the unicity problem once and for all.
Given the ubiquity of smartphones, accessing asset attributes by scanning a visual code (QR code, bar code, etc.) is very easy and convenient. Far more information is accessible this way, without mental gymnastics.
This is how the Trak platform solves the ID generation issue for asset tracking (and other process tracking use cases). We generate global IDs beforehand, and encode them as QR codes. Contact us today for a free trial.
