As software experts in the field of Industry 4.0, we were rather surprised by this realization when we took part in a lean production training. Why? When it comes to introducing Industry 4.0 solutions, decision-makers and experts from both manufacturing and IT have certain expectations. In general, they expect that the mere possibility of storing and analyzing large quantities of data – along with expanded automation and monitoring functionalities – will deliver increased manufacturing efficiency, transparency, and flexibility as well as fault prevention. On top of this, Industry 4.0 solutions can ideally analyze data in near real time. Right across the value stream from suppliers through to customers.
Why might such a theoretically powerful combination fail?
I’ll use a few examples to show why these expectations can’t always be met.
Industry 4.0 projects as “autonomous milk runs” give rise to self-organizing, flexible materials transportation systems for supplying production stations – in place of cyclical material supply, which is defined in the Bosch Production System (BPS) as an element that helps implement lean production.
At first glance, autonomous milk runs contradict a series of BPS principles:
- They reduce the transparency of internal logistics streams.
- The resulting lack of standardized materials transportation can lead to a situation in which deviations from plan and hence faults are no longer immediately detected.
- A switch to autonomous milk runs turns the calculation of maximum replenishment times into a complex or even impossible task.
- Further optimization becomes more difficult and/or time consuming.
Value streams in manufacturing are generally so complex that a change at one point (e.g. software-supported optimization of cycle times) can have a major impact all along the production line.
In the case of software-based monitoring of cycle times, there can also be employment law consequences, leading employees to reject what they see as pervasive surveillance.
And finally, if data analytics and other software doesn’t provide experts with the reasons for certain outcomes – which would put them in a position to exercise smart control – then the principle of “responsibility” is compromised.
The principles and elements of lean production that stand behind the points I’ve mentioned are essential components of the continuous improvement process (CIP) in production. In turn, continuous improvement is the way to achieve sustainably waste-free processes.
Lean production and Industry 4.0 can synergize only if
- production systems have been suitably designed (“first process maturity, then Industry 4.0 ”) and
- IT experts understand the complexity of the production systems and can offer appropriate advice (“combine manufacturing & IT expertise”).
This is why Bosch places such emphasis on closely interweaving the BPS and software experts, a measure which also ensures that the kinds of mistakes mentioned above are avoided. The software engineers are given the tools to understand and resolve any conflicting principles that may arise. They are encouraged to implement Industry 4.0 in the Bosch production sites in accordance with BPS principles and elements – with the goal of effectively combining lean production and Industry 4.0 .
Training courses exist for software experts who are helping to implement Industry 4.0 solutions in Bosch’s 250 plants around the globe that also use BPS. I myself was able to take part in one of these training courses and work on a project to reorganize a simulated product line according to consumption control principles in three stages. This was done using classic lean production approaches: we implemented the pull principle with kanban cards, optimized cyclical material supply, created transparency of goals, etc.
In our example, the software experts have learned and observed the guidelines in order to maintain the requisite transparency, personal responsibility, fault prevention, and standardization when new Industry 4.0 solutions are created and introduced. This includes the following measures:
- Automatic software-controlled rescheduling requires confirmation by the relevant production expert. As part of this measure, the software notifies the expert with information about the reason for the required rescheduling.
- Replacement times must be guaranteed despite automation.
- The production processes must already be mature and understood before Industry 4.0 is implemented.
- Gemba walks remain a key part of improvement work. You can only get a proper overview on the shop floor. Software must support this.
- Transparency must be ensured at all points of the changeover when introducing the new software. This may involve running manual and digital systems alongside each other for a defined period of time.
- The digital data must be prepared such that the production worker is able to immediately identify deviations and recognize the causes.
- Some data analytics expertise must also be available on the shop floor. This can mean that the data science experts in the Industry 4.0 project should initially accompany the acquisition of this expertise with respective trainings for the production site experts.
What has your experience been? How well do lean production & Industry 4.0 complement each other?