paulgrehan
commented
1 year ago This is an interesting perspective as we struggle with data integrity versus data provenance, the "VW" issue. With asset certification we have relatively mature models to test/verify, pair this with VCs for lifecycle authenticity would capture both aspects potentially. Keen to help/contribute to this, it's not exactly what we are looking at but has the same core issues... 👍
Description: To date much of the carbon reporting in the energy industry is based on modeled data. Model's have the many risks built into them and can be easily manipulated. Today asset data is collected manually or via spreadsheets and entered or uploaded into models. This has the double risk of fraud opportunities and significant cost due to the manual processes. As reporting requirements increase the current processes are no longer viable.
Opportunity By packaging sensor data into a digital credential linked to the asset DID, and owning organization DID the risk of data on motion attacks would be significantly reduced.
There is a potential conflicting requirement on how the data comes off the sensor. Sensors in many industrial environments are connected by network transports other than HTTP, and these transports can impose severe constraints on packet size and transmission rates. W3C VCs in JSON-LD format are very large and not feasible for transmission in these environments, however the linked data model can provide real value in post processing of data.