1. Improving efficiency of industrial production
A lot of work is being done to improve production efficiency with digital, IOT, data and other new-ish technologies such as digital twins and cobots. Basically every company from the chemicals industry to automotive is working on sensoring devices and processes, deploying IOT platforms, as well as working with data and AI to improve efficiency of their processes.
Yet, the event polling figures show that currently less than 50% of industrial companies have a comprehensive data strategy in place, and less than half are investing in the right technology to exploit data they have.
At the moment industrial companies at the event were frequently expressing that their production data is still spotty and difficult to interpret. The biggest improvements can sometimes be made in areas where sensor data just does not apply, at least yet. One of these is the example of resetting production and process lines for a different product.
Few industry examples:
BMW is using data and AI fairly extensively to make production more efficient and of higher quality. For one, they have nice results with simple AI image recognition tools automating simple and repetitive tasks, such as car type verification and database entry, real-time measurement of fast-moving materials, as well as quality assurance at car body paint shop.
Hugo Boss, in the traditionally low tech clothing industry, is seeing four layers in their Industry 4.0 from manufacturing perspective: people, products, processes and machines. They've used digital for training their people using what they've learned from the gaming world. The results have been good, with employees taking more inititative in organising work and improving productivity.
In terms of artificial intelligence, Hugo Boss has no shortage of ideas. However, what they have found out is that while they get more insights from processes and can in some ways improve their life, monetizing data and AI has been hard. There have been a couple successes, though. AI-driven production planning is one. Hugo Boss found that their AI solution is better than their experienced employees in planning certain production activities, such as cutting. Another success case comes from human resources. They are predicting employee demand with AI, and increasing productivity via AI-driven employee rotation.
Airbus again is an extensive user of large scale robots in airplane manufacturing, with a focus on moving towards commercial use of cobots (collaborative mobile robots) within months.
Solenis, a specialty chemicals company is working to sensor customer processes and improving process efficiency based on the collected data.
Beyond the manufacturers themeselves we are seeing a lot of platform+consultancy companies providing platforms, tools and advisory on improving production processes. Examples include Elisa Smart Factory solution and German Edge, both working across manufacturing and process industries.
2. Going up in value chain to services, outcomes and autonomy
Creating intelligent products, new services and business models seems like a tougher task for business compared to improving the efficiency of existing production at industrial companies. Things are happening and going forward, it just looks tougher and less clear for the participating industrial companies.
The talk is of industrial businesses moving from selling products to selling services, with business models going towards pay per use and pay for quality and eventually towards pay per outcome and autonomous services.
The feeling you get is that there is still a lot of work to do in moving companies from products to services, while the outcome and autonomy phases are even further away for industrial companies. Yes, some very advanced (near)autonomous products exist, released e.g. by Airbus and a few others.
At the same time there are new players that are seeing a business opportunity. The new competition is attracted by the new positions and new revenues streams that are up for grabs. One interesting example of a new player is Relayr. Acquired early on (at valuation of 300m Euros) by a sizable insurance company, Relayr is an Industrial Internet technology company positioning themselves as the enabler of as-a-service or pay-per-outcome business models for industrial companies. Although tech company, they have significant expertise on as-a-service commercial models and financing. Industrial companies should take note.
Related to the challenge of moving forward toward new services, business models, strategic control points, and continuing to "own" the customer, Karim Pourak, CEO of Process Miner, noted in a panel discussion that short term value continues to be damn difficult to present for the renewal investments. This is an all too well-known big problem, that also includes a risk of delying the investments and investing too little.
Another point was made during a panel discussion regarding budgetless CDOs in industrial companies. She argued that for CDOs to actually succeed in their work, they need to be given clear power and budget to really push the digital business transformation forward within organisation. Currently, this is not the case. What is the CDOs role, then?
Note: I think that most budget should go to business units, while CDOs need to have some seed funding to encourage and push experimenting forward.
Regarding the approach to Industry 4.0, this is what was said:
- Everything that uses electricity will be connected online
- Everything that can be automated and robotized will be
- Everything that can be digitalized will be
- Everwhere that AI can be utilized, it will
Over time, strong AI will come
Six key points:
- Don't just focus on the factory floor. You need to look at the whole value chain and value creation. That's where the money is.
- Do not drown in your data lake, but focus on business value and work back from there.
- Build cross-functional themes and get out of silos. Digital teams need to work within businesses, not separately.
- Find digital and data value in steps, incrementally. The big bang approach will fail. Idea is king, execution is King Kong.
- Start where data is available. Find existing data set, use existing point solutions.
Build organisational capabilities down to cultural change level. Yes, culture eats strategy for breakfast.
That was day one! Some impressive demonstrations by IBM still going on.
To read more about our thoughts on the future of the industrial sector, please download our newest point of view paper here.