By Geir Rune Flåten, Chief Solutions Officer
Timely, accurate and trusted data has never been more important than it is now during this pandemic.
Every day we read about new candidates for COVID 19 vaccines and in the race of the pharmaceutical companies, data is a pre-requisite. Not only for FDA audits and approval, but also for a fast scale-up from lab to pilot to production.
Data is at the core of the value chain from development to production to business, and to derive the most value, data should be shared and actively used across the organization when solving problems with compliance, development, quality control and business development. The keyword is data transparency.
Not so long ago the capability of reliable measuring, fast analytics and automated use of actionable insights was a challenge. To me it now seems the barrier for getting more value out of data is much more about organisational structure and way of thinking. The technical barriers are gone.
Data transparency is an important aspiration in terms of industrial data. I am aware that the term is usually associated to privacy issues when companies and institutions are digitally entangled with individuals and companies. In the industrial space we are luckily operating with data outside the scope of GDPR, but nevertheless data transparency is a desirable state and not a given.
Let me give an example. Decisions about sensors, equipment, software and data formats in a given manufacturing process may be taken locally with a “narrow” perspective of getting the job done. Narrow because the data collected may be useful in other branches of the organisation and in other phases in the product lifecycle. In a global company with 10 or 100 manufacturing sites it is evidently valuable to share knowledge and not make the same mistakes, when the lessons already should be learned.
– When spectroscopy data are translated into chemistry it gives valuable information about materials, products and processes that should be used and re-used for the benefit of the company. The cost of industrial data is significantly higher than consumer data, so to me it is common business sense to optimise the value from industrial data.
I am not advocating for companies to pour all raw data in a data lake and wait passively (and aggressively) for meaning to evaporate from the swamps of data. It will probably not happen. The key is a radical openness about lessons learned, conclusions made and connections identified for value to be achieved elsewhere. This is key to further automation and taking steps to realise the visions associated to Pharma 4.0.
When spectroscopy data are translated into chemistry it gives valuable information about materials, products and processes that should be used and re-used for the benefit of the company. The cost of industrial data is significantly higher than consumer data, so to me it is common business sense to optimise the value from industrial data. Data capture through specialised sensors is often expensive, which should give us even more reason to secure the data transparency that potentially multiply the value of insights produced.
If only the “narrow” goal of the short-term purpose is present in the decision process, we lose sight of the value of the analytical outcome in other disciplines and at a larger scale. This makes silos within the organisation, the most important obstacle for data transparency of industrial data.
Another obstacle is proprietary lockdown of data from vendors of sensors, equipment and software. To obtain data transparency you must evaluate vendors for true openness in terms of standards and interfaces and willingness to contribute to a wider community of people, technology and data. One baseline could be for your vendors to support OPC UA and in the bigger picture walk the talk of openness.
What we do not want in terms of Industry or Pharma 4.0 is isolated islands of data fenced by proprietary technology or walls consisting of pricing models. This is a pre-requisite for more automation and more data driven development and manufacturing. Quite parallel to the terminology in business intelligence and analytics: In the industrial data space we also need “one version of the truth” and true “data democratisation”.
Talk to an expert
Get in touch if you have questions about our products, platform, how to get started or how best to address your analytics needs.