How Musk came from the shop of the future to the assembly in the tent

In the continuation of the topic of how Musk still wriggled out and produced the necessary 5,000 Model 3 for the week. But how? Not really with the help of the most widely advertised fantastic robotic line of the 22nd century, where there are absolutely no people and only robots are scurrying back to here as they showed us in the commercials?

Let’s see with the help of Stanislav Bezgin what Elon Musk’s attempt to carry out a revolution in automotive industry resulted in, especially since it was exactly one year since the launch of the most modern assembly line in the world at a factory in Fremont.

A year ago, with the pomp and hip, the promised alien dreadnought was discovered – a new assembly line, built and designed by troika in sketches and fantasies by the humanities. Investments in the amount of two billion dollars were to ensure the production of five thousand cars a week, and this is the amount of output that the factory was supposed to reach by the end of December 2017. For a better understanding, the situation must be taken into account that the factory had to operate in “normal mode “, As are all the usual automotive production in the world, that is, in two shifts and five days a week. And, therefore, the design capacity of the factory was 2,500 Model 3 machines in one week’s shift.

The factory opened … and, it is expected, immediately rose stake. The factory did not work from the word at all.

From a recent article in the WSJ:

“Tesla installed 1,028 robots at a factory in Fremont, about one third of which were uniquely suspended upside down so that the company could squeeze more into the factory’s space.”

Uniquely suspended robots and usually installed or not worked, or did everything wrong, automatic lines did not provide spare parts, there were always problems at each site. Musk called it a “production hel,” and moved to sleep on the floor of the factory. This helped a little, because for the first quarter of the factory the factory produced only two or three hundred cars, and those were assembled manually on some side platforms.

Talking about tuning and optimization, it took six months of effort to achieve at least some output. By the appointed time of full launch, the company managed to reach only the output volume of about 200-250 cars per week, that is, about 10% of the planned capacity.

To avoid the impression that the saboteurs who entered the personnel department deliberately conducted a negative selection of engineers only for the plant in Fremont, let’s look at the other separate production divisions of the company:

Simultaneously with the assembly line for model 3, the battery assembly line for this machine started working. This is the one that was supposed to produce batteries faster than a machine gun. Robots either did not work, or just spoiled the raw materials. The line was not even close enough to ensure the release of batteries in the right quantities. Moreover, despite the fact that the main assembly line in Fremont did not work properly, the line of batteries could not periodically provide components for these even poor rates of work. After nine months of “adjustments” and “optimizations,” this resulted in the dispatch in May of planes (!!!) from Germany of a new assembly line. It remains to be hoped that this time the German engineers did it not on the basis of the fantasies of the humanities;

Gigaffactory producing powders (home batteries). According to the information of the dismissed employees of the solar unit: “Delay in deliveries of power stations has always been from six to nine months, with declining sales volumes. Because of this, we tried not to offer customers a powerpan. “;

Gigaffactory producing solar roofs. The first customer received the products with a delay of one and a half to two years, and until now the products are produced in minimal quantities.
So, the ability to create problems on an equal footing and then solve them for a year is a common feature of any production unit of the company. Meanwhile, in Fremont, Musk, radiating rays of optimism, continued to talk about temporary difficulties and the continued “tuning” of the work of the line. Three months later, by spring, it was possible to bring the plant to a capacity of 400-500 vehicles per week.

And here it became clear that for the constructed superautomated factory, similar to the alien dreadnought with shining and sparkling robots, this is the actual and theoretical limit. You can get the factory to work four shifts a week and get about two thousand cars and this was done in the last week of March. But this is a ceiling, above which no further “settings” and “optimizations” can break through.

We can say that at that moment the fifty-billion-year-shining project “Tesla Corporation” burst, Elon Musk burst out and splashed and all those around him. That the most interesting majority of people did not even understand the importance of this moment. But this is well understood by the top managers of the company, and it was then that their team competition “Hurry to escape – in order to get rid of the shares” began. Within a few months, the company was left almost all the key figures:

– John McNeill is the president of [global sales] units;
– Eric Branderis, Chief Accountant and Vice President;
– Susan Repo, treasurer of the company and vice president for finance;
– Matthew Schwal, Senior Executive Director;
– Doug Field, Chief Engineer and Vice President of Production (although his disappearance was concealed until the beginning of the third quarter).

As of today, in the company of significant top managers, only Elon Musk himself, his financial director, Deepak Ahyudzha and technical director of DB Strobel, remained, in fact.

In March-April, realizing that fantasies about robotic assembly of cars turned out to be premature, Musk reported that he was wrong, accused Dag Field of production problems, dismissed him as vice-president and reported that he would remove the robots, replace them with people and will recruit 400 people a week, until it provides 24-hour operation of the factory in 24/7 mode.

“Excessive automation was a mistake, to be more precise, my mistake. People are undervalued ”

The problem is that it is impossible to take a nonworking, incorrectly designed automatic line and throw some of the robots out of it, to make a high-performance modern assembly line combining automatic, semi-automatic and manual assembly. Such lines are designed and mounted as a mechanism for a specific model (or several similar models) and on the knee it can not be done.

Since the beginning of the second quarter, a struggle has begun to achieve the opportunity to show investors the production of four to five thousand cars a week at the end of June.

It is very characteristic of how the Musk provided its workforce needs, some information about it was given by the New York Times journalists:

“Workers experience pressure with the demand for speed of release. In an interview outside the factory, several people said that they worked in shifts for 10 and 12 hours, sometimes six days a week. They report that the turnover rate among the collectors is high, and that sometimes the chiefs get up to the conveyor during extended shifts.

One of the difficulties that workers see is the rapid influx of new workers. The company aims to hire about 400 employees a week to speed up the production of model 3. After the most recent Tesla report on incomes in early May, Mr. Musk said he hopes he will eventually have three shifts per day, essentially , working on the assembly line around the clock.

“Everyone I talk to here has only been here for two weeks, a month, and these people do not stay long,” said Jonathan Galescu, a body repair technician at Model X, who worked for the plant for four years.

Mr. Harbor, a manufacturing expert, says that automakers usually train new workers for several weeks before allowing them to work on the conveyor belt. Attraction of a large number of new employees can worsen the quality, because they may not perform their work properly or fail to notice when problems arise.

New employees at the Fremont plant receive three days of training before commencing work on the assembly line. This includes a day of computerized virtual training to safely perform their work and a day of training in the area to which they will be assigned. ”

It’s amazing that with such personnel, with a three-day education and two-week experience, Musk was able to produce something at least. Nevertheless, in April-May, Musk was able to reach a productivity of 650-800 cars per week. With the operation of two or three shifts of semi-trained workers, it allowed to produce about two thousand cars a week. With such average rates, the plant worked in the second quarter except for the last week.

That is, when working twenty-four hours a day and all week it would give a maximum volume of 3500-3700 cars per week. Again, this is almost the limit for the factory in Fremont, since there is simply not enough room inside to simply re-engineer and place a normal assembly line.

In May, this became obvious to the Musk, so the creative team was tasked at any cost to depict at the end of the month the output to reach a volume of five thousand pieces. Someone, perhaps Musk himself, suggested an elegant solution – on a parking lot near the factory, to put a huge tent, the size of two football fields, and in it to place an assembly line assembled from all scrap and illiquid lying around the company’s warehouses.

This solution had two advantages, first, it could be implemented quickly and meet the deadline until the last week of the quarter, and secondly, there were no robots in the clean place that interfered with the normal work of the collectors.

Result Musk really liked, a new assembly line on Twitter, he called “sweet” (sweet). In an interview with the New York Times, Musk said:

“The line [in the tent] does everything that other assembly lines do, but with fewer people, reduces labor costs and significantly increases uptime. Our cost per unit is lower on this line than on other lines, and we see a higher initial quality. ”

To illustrate what is at issue, I will give photos of the process of assembling cars in the last week of the previous quarter.


This is a bunch of Mexicans (and other Americans), cartons, forklifts, primitive conveyors, hand screwdriver (nut) vertices – this is an assembly line that surpasses the quality, speed and number of people, that strange hybrid complex that turned inside the building on the basis of half-working robots and chaotically wedged manual assembly posts?

Obviously, it is impossible to install automatic, semiautomatic or simply complex and expensive equipment in open from the ends of the tent structure, as a combination of dust, sand, moisture (condensation from day / night temperature differences is guaranteed), heat, vibration and high loads will kill such equipment for half a year. In fact, in GigaTent there is only the simplest equipment for lifting and moving loads on the conveyor. I think that even Henry Ford, if he were released from hell on a business trip, would not find any innovations in the assembly line located in GigaTent.

And, of course, a line of random elements assembled in two weeks, on which semi-trained employees work, is not capable of giving performance at the level of industry standards, and it works about 5-10 times slower, and yet Musk claims that it’s all the same better than the “old lines” of extraterrestrial dreadnought, and in this case, probably, he is not much mistaken.

What is the goal of the Musk as a result of a week-long race to five thousand cars? Has Tesla become a “real auto company” after that?

Well, first of all, it was, of course, more of a show before the stock exchange than a real production task. Musk used the term “factory gate” when submitting information about a record. According to the information obtained by BisinessInsider, the difference is that at the same time machines are added, but produced on the conveyor belt, at the time the period starts. How many such assembled cars were prepared at the beginning of the assault week the story is still silent;

On June 26, Musk’s order was forbidden to carry out a number of critical test operations of the brake system (this can be seen from internal documents that were placed at BusinessInsider’s disposal), which slowed down the process of “factoring”. Who cares about quality and safety, when the only task is to deceive investors. After getting this information into the press, the shares, by the way, fell by 4-5%;
Normal technological breaks for people were canceled or reduced;
Judging by the low volume in the previous weeks, the factory devoted part of the time to harvesting pre-assembled units and accumulating scarce components;
Well, in fact, Musk forced the factory to work out another five hours in excess of a week’s time, in order to bring the account to five thousand cars.
That is, in spite of any efforts, the factory is not yet technologically ready to produce five thousand pieces a week, and, moreover, to a stable output of high-quality products in the volume of about 21,000 cars per month.

By the way, about a thousand machines were assembled in GigaTent. That is, today the peak performance of the plant itself is 3600-3700 pieces per month. (This is if it is normal to withstand the technological process).

Of course, the factory is not yet ready and practically, since the Musk still does not have the necessary labor force. Despite the fact that “people are put to the machine” at all without training and, apparently, without careful preliminary selection, Musk still has workers for no more than two or two and a half standard shifts. In the June assault week, to ensure the maximum loading of the plant for seven days:

– Managers and employees of non-core units of Tesla were sent to the conveyor front;
– Many people worked for twelve hours, six days a week;
– collectors of old models were recruited from time to time (although not too many, since they managed to “factor in gating” slightly less than two thousand old models);

To give people a rest after this week’s hard labor, Tesla stopped the plant for four days. Actually, if each time after such an exemplary performance, stop for four days the plant, then the average productivity per week will be less than four thousand pieces.

Actually, without forcing people unnecessarily, the Musk in July-August can still work only at a rate of two to three thousand pieces a week. He has no people, and vaguely suspect that the more he forces people to work for twelve hours a day, the higher his turnover will be. The situation is aggravated for him by the fact that now it’s summer, the season of holidays, and skilled workers (with experience) demanding to give them the opportunity to relax with their family and children.

And we need to understand that the high-speed construction of GigaTenta was necessary only to deceive “investors”, since for production purposes the tent was absolutely unnecessary. The fact is that in June and in July the Musk is not limited by assembly capacities, but only by the number of people it can put on these capacities. Until the next assault week, a tent with Mexicans will take a minimal part in the assembly of cars, well, or, alternatively, if it is really better and more convenient than domestic lines, then production will be partially discontinued inside the plant.

Apparently, in the period until the end of the quarter, Musk will be able to recruit, train and retain enough people to sustainably produce three and a half or four thousand cars a week, and exemplarily produce six thousand cars a week. By the way, the next assault week is scheduled for the end of August. But you can imagine at what price this achievement will be achieved. I do not think that the California labor laws are allowed to force workers to work at night and overtime, without compensating for this with higher wages.

At the same time, further dismantling of robots will take place, and further training of workers. This will increase the overall productivity by 10-20%, but it will not solve the issue of quality, speed and cost price of the plant.

The musk faces an insoluble problem:

– He is not able to sell a lot of expensive version of the car;
– It is not able to produce a lot of cheap version of the machine;
– He is not able to maintain a network of fuel stations, sales and service centers, selling a little expensive version of the machine.

Without the key element of the development of the corporation, without a robotic car assembly enterprise that allows to bypass competitors at the cost, quality and speed of assembling cars, the whole development strategy turns into a mixture of good religious sermons, naive fantasies about the future, extortion of state subsidies and frank deception of sponsors who are still thinking, that they are investors.

Here there is a lot of reasoning on this subject from Stanislav Bezgin (also known as Toyama Tokanava)

Back to top button