Very good conversations on April 24th and 25th, 2018 at Dechema´s forum „Big Data Analytics in the Process Industry“ in Frankfurt, with several experts from the chemical industry and technology vendors in the area of big data / predictive analytics and artificial intelligence (AI). Therefore, first of all, a big „thanks“ to Dr. Björn Mathes and the Dechema team for great organization & moderation!
What we discussed:
- Is AI mature enough for the industry, and what concrete applications can it have?
- What technologies are available for big data / predictive analytics, and which are major differences?
- How to apply big data / predictive analytics to improve chemical innovation & operations?
- How can IoT and real-time data generate value for a corporation?
- How will quantum-computing impact the digitalization roadmap, and where does the research stand today?
What we did not discuss – and would be interesting to do in future occasions:
- How can big data analytics be integrated in the overall context of a chemical company´s digital transformation?
- What are the implications of big-data analytics approaches in terms of data / cybersecurity, and what critical decisions are required from chemical companies?
- What skills will be required in future to drive a big-data „enabled“ chemical company, and – more importantly – how to pragmatically acquire and integrate them?
- How to manage the overall digital transformation process – and especially – cultural gap between technology experts and business experts?
And here my personal takeaways – thereunder some confirmations, but also some quite interesting new learnings:
- AI and predictive analytics technologies are broadly available at several reliable vendors and ripe for use in the chemical industry. We discussed several concrete applications already realized: from R&D to sales / forecasting down to operations. For these we saw use caeses & business cases and, no doubt, technologies are there ready for use! I especially liked the cases brought by Jasper Rutten of Huntsman, Dr. Safa Kutup Kurt of Merck, and Mirko Schnitzler of Infotopics.
- In particular, innovation and process development & improvement are two areas that can benefit a great deal from them. After a drop in patents filed in the 2000´s, innovation activity picked up again in chemicals after 2010, showing increased attention to innovation as major driver of growth. Here we saw great opportunities to apply AI and big data approaches to expedite time to market (which will become the critical factor to succeed in many markets) and reduce costs, one among other the use cases shown by Dr. Michael Sokolov of DataHow.
- There is a lot of potential in applying AI to new material discovery. I have been very impressed by the presentation of Douglas Ramsey and the approach of Citrine Informatics – an AI-powered data platform for materials discovery – and the support it can offer to private companies to boost their development while maintaining their IP safe.
- The old problem of data quality input is still mostly unsolved. Technology, alone, is not yet able to sort good data from bad data. We used to say „garbage in, garbage out“, and it seems that this old adage still holds up to these days. Whether you are using established statistical tools like R or more modern devices like JMP from SAS (which is quite powerful by the way – thanks to Mahmoud Hammoud to show us), you´ll still need someone to sieve your data before using with the software.
- All in all, technology and business still have some gaps to bridge before going hand in hand. There is a tendency of Technology providers to try to fit their solution to your problem rather than understanding your problem. The way between technology and concrete, working application is still long and not charted well. Also, there are a lot of providers with great tools out there: technology is clearly available, the challenge is to choose the tool that is good for you and deploying it productively and broadly in your organization.
- The transformational aspects of introducing digital technologies are often played down or not addressed at all – as if digital transformations were just „plug & play“. I perceived some lack of attention in the convesations about transformational aspects. Now: the chemical industry has 300+ year of history and most chemical companies have very established, traditional, often conservative corporate cultures, value sets, and organizational structures, not to mention a fairly senior employee population. While we discussed the provocative vision of having „digital chemists“ (AI systems trained to do specific tasks around product innovation in chemicals; multi-agent cognitive digital twin architectural principals), I was just playing in my mind the movie of the range of likely scared reactions of the average chemical lab technician I met so far. Good discussions here with Burghardt Schmidt and Andre Bosman of ACS that brought in their very specific perspective as vendor-independent technology integrators.
- As a consequence, there is an urgent need for digital change agents and business-to-technology translators: traditional IT departments may evolve partly in this direction. This is still an open point for debate as we heard in one of the presentations from a major German chemical player. Traditional IT departments are looking for new ways to add value to their business and taking a more active role to push the transformation, but these efforts seem scattered and not durable / established. Often companies recur to external consultants or experts – who, however – are typically engaged for one specific project and leave thereafter, without really changing the mindset and the people.
- Last but not least: the chemical industry is way behind in terms of digital adoption versus many B2C counterparts. There was a broad consensus in some side conversations that the chemical industry did not feel major pressures to aggressively go digital so far and that B2C businesses are light years ahead, eg. leveraging consumer data & sentiment from social media to take demand, supply, and quality decisions. I believe here that in the foreseeable future the voice of the (end) customer and her quest for quality / speed / agility will become so loud that it will be felt all along the chain from B2C to the most far away B2B businesses: the shorter the chain, and the more interconnected the players, the sooner this will happen.
I definitely recommend to anyone who is passionate about the chemical industry to participate to the Dechema forum next year and look forward to discuss new experiences & learnings then.
About the author
Lorenzo Formiconi is a partner at execon partners. He is an expert in operations management and digital transformation, working with clients in the chemical industry since 20 years.
For further information contact Lorenzo Formiconi via email or phone.