GDPR frames everything in the context of a person's data. There is no "person_id" or similar field in these data models. That isn't the purpose of the data, it would be expensive to extract it, and then it would create obvious liability under GDPR. This makes the idea of finding a person's data expensive -- brute-force search on huge data volumes.
Compounding this, these data systems are often operational and some of the data may be in situ at the edge because it is too large to move all of it. The power and compute budget may not exist to find a person using brute force.
AFAICT, current best practice is to maintain a polite fiction that people aren't being tracked because that is not the intent. No one thinks that would stand up to serious legal scrutiny though. If the regulators come after you then plead best effort based on the technical infeasibility of doing anything else.
Forget tracking workers' movements and stuff like that because that's even more complicated (the data is tied to a person, but only in their capacity as an employee and not as a private individual).
Focus on a case like a cluster of sensors attached to various equipment powered by electric motors, or using RFID to detect when a pallet enters the warehouse. Let's say that all goes to a cloud platform and I store it, I build a bunch of derived analytics stuff from it, and I send it up to Anthropic (with no-train-on-me-pls contract clause) for my cool new AI insights engine.
Does GDPR apply at all to that? I would have assumed it doesn't have any relevance whatsoever, but you're implying that it does. Or are you specifically talking about the case when individual employees are the data collection subjects, like a fleet management platform with a telematics component?
--- start quote ---
Taking into account the state of the art, the cost of implementation and the nature, scope, context and purposes of processing as well as the risks of varying likelihood and severity for rights and freedoms of natural persons posed by the processing, the controller shall, both at the time of the determination of the means for processing and at the time of the processing itself, implement appropriate technical and organisational measures, such as pseudonymisation, which are designed to implement data-protection principles, such as data minimisation, in an effective manner and to integrate the necessary safeguards into the processing in order to meet the requirements of this Regulation and protect the rights of data subjects.
The controller shall implement appropriate technical and organisational measures for ensuring that, by default, only personal data which are necessary for each specific purpose of the processing are processed. That obligation applies to the amount of personal data collected, the extent of their processing, the period of their storage and their accessibility. 3In particular, such measures shall ensure that by default personal data are not made accessible without the individual’s intervention to an indefinite number of natural persons.
An approved certification mechanism pursuant to Article 42 may be used as an element to demonstrate compliance with the requirements set out in paragraphs 1 and 2 of this Article.
--- end quote ---
IANAL, but this basically covers all your bases, together with https://gdpr-info.eu/art-32-gdpr/
Unless, of course, your industrial-scale data collection actually collects significantly more data than you let on, and extraction of personal data is not as hard as you make it sound
The GDPR is there to protect your personal/sensitive data, or data that can personally identify you. If has nothing whatsoever to do with data capture from industrial machinary.
I remain astounded how ignorant some people are of basic GDPR principle: protecting your _personal_ data.
How is this not your personal data?
Exploitation of these types of data sources has been demonstrated for 15+ years at this point. Abuse is often impractical for technical reasons but GDPR doesn't give you free pass on collecting personal data just because you aren't using it like personal data.
Many systems were not explicitly designed for surveillance, and are. Because many systems collect too much data to begin with.
Hence the problem: people who collect too much data claim that GDPR is complicated, complex, convoluted, impossible to comply with... instead of changing what data they collect, and how.
Additionally, people confuse the complexity of human endeavours with the complexity of the law. GDPR itself is neither complex nor complicated. It doesn't try to carve out exceptions, rules, and regulations for every possible activity humans may attempt. Then it would become impossible to understand or comply with.
As is, it has enough carveouts for industries which require more data than strictly necessary, called "legitimate interest" (which still doesn't allow you to just use this data willy-nilly). E.g. banks collect significantly more data about customers than strictly necessary (because KYC, fraud, security etc.), and store that data for significantly longer amount of time than allowed by privacy-related laws (because they are governed by bank laws of respective countries). It doesn'tmean they can sell that data or spy on users.
Same here. It's not on the law to tell you exactly how to operate your "industrial-scale operation". It's on you to fix your shit, stop collecting more data than necessary, have data protection in place, delete data after a reasonable time, anonymize data etc.