After the Giga Factory, Tesla introduces the Giga Press, a giant press made in Italy produced by IDRA Group. The role of the Giga Press is immediately stated: thanks to its imposing dimensions and a force between 55,000 and 61,000 kilonewtons it is able to print a huge portion of the Model Y body, reducing the pieces from 70 to 1 to create it and replacing the work of 300 robots (source: https://insideevs.it/news/485909/giga-press-italiane-tesla-video-ufficiale/).
In these terms, it would seem an easy analysis: higher productivity, fewer parts, fewer robots, less logistics and maintenance costs due to the reduction of the number of robots required, but also higher quality of the car body. However, the analysis is not as trivial as it might seem.
For Lean fans, for example, the Giga Press represents a “Monument“. As stated in the famous book “Lean Thinking”, lean thinkers define a monument as any machine that is too big to move and whose dimensions require batch production.
Therefore, given that continuous improvement and changing process requirements involve the constant movement of machines, monuments are another form of muda.
The Pratt & Whitney case study
The author reports the case study of Pratt & Whitney, a leader in the production of aircraft engines. The monument we are talking about was a gigantic complex of 12 grinding centers valued at $ 80 million, custom built in 1988 by Germany’s Hauni-Blohm, when Pratt tried to make a technological leap ahead of its competitors. This center had to replace a series of 9 grinding machines for a total process time of 84 minutes. On the other hand, the new system was able to grind a blade in just 3 minutes (-96%) and the centers were loaded and unloaded with the aid of robots and AGVs, without the need for direct or indirect work.
One of the fundamental problems was linked to the setup time required by the new plant: about 8 hours.
However, Pratt needed to produce small batches of many different types and the long setup times made it impossible to do so and instead required the production of large batches of each type of blade.
Moreover, many general workers had to be replaced by about 22 specialized technicians, a number not much lower than that of the old manual system.
Eventually this complex and expensive machining center was replaced by a “chaku-chaku” (charge-charge) system with eight simple grinding machines with quick-change systems (set-up <2 min) and an employee able to move components from one machine to another and to measure parts to check their quality.
Processing times increased from 3 to 75 minutes, but downtime was reduced by more than 99%, space required reduced by 60%, the average batch increased from 1640 to 15, the total production cost was halved compared to an initial investment of $ 1.7 million (instead of $ 80).
Tesla’s Giga Press
Going back Tesla’s Giga Press, will it be a pay-off investment?
We are confident that Tesla engineers did their quantitative analyses before making such a major investment. However, let’s try to make some considerations in this regard, based on the information found on Google:
- The investment considerably reduces the number of parts to be pressed, thus reducing the associated management costs
- The number of robots (about 300) required to weld the 70 components is reduced
- This leads to a reduction in the footprint required for the warehouse and robots, no costs for testing the welded joints and a general reduction in the cost of production, but at the same time an improvement in the quality of the car body.
- Cost reduction is probably one of Tesla’s main drivers, which aims at the mass diffusion of the electric car, making it increasingly accessible
- We do not know if the Giga Press will be dedicated to a single mold or if it can be retooled and in how long. Certainly, in this sense Tesla will pay in terms of production flexibility and in the ability to react quickly to changes in demand
- We do not know the reliability level of the Giga Press, but probably given its complexity it is reasonable to think that it is inferior to that of a robot. But compared to 300 robots?
- We do not know how many maintenance workers there will be and what specific skills are required: they will certainly be more specialized than those required for welding robots
- Compared to the Pratt case study, Tesla operates in a different market: higher volumes and minor product variations. However, the trend towards customized manufacturing will increasingly be a key competitive factor for the years to come
- Nevertheless, the creation of standard platforms and the creation of product families in order to reuse the same components will play a fundamental role in the general cost / benefit analysis
From these considerations, we can deduce that:
- There is no concept better than another, if not according to the specific requirements, objectives and system constraints
- Small and flexible isn’t always better than big and rigid, and viceversa isn’t always true
- A highly efficient system with a different approach is achieved at the BMW plant in Munich
- It is necessary to have a medium-long term vision of one’s product portfolio in relation to one’s production system, and act accordingly
- The decision should be made as far as possible in quantitative terms, with real data
- In this sense, a discrete event simulation could provide a fundamental help in simulating the different scenarios according to the technological solutions indicated.
What do you think? Which solution do you prefer?