-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Tesla Gigafactory
- Thread starter domenick
- Start date
You are probably right. Images from May 18th and 20th on Planet.com indicate another section being "re-roofed". does not bode well for quality of construction if that is the case.
JRP3
Hyperactive Member
May 4 flyover shows a larger black area Latest Tesla Gigafactory flyover video reveals newly built sections [May, 2017]
AZ Desert Driver
Rare combination
Rather than assume a problem, perhaps they need to install some very tall equipment and need roof access to accomplish this new feature.
Thomas Edison
Active Member
The science program Impossible Engineering aired tonight, and showed the gigafactory with solar panels covering the entire first section.
AZ Desert Driver
Rare combination
Any way to slow down, pause the progression of these photos?
Thomas Edison
Active Member
FlatSix911
Porsche 918 Hybrid
Tesla had a big party last summer along with tours of the GF for referral winners
winfPanasonic post: 2142569 said:
The only official number is 477 employed by Tesla and Panasonic at the end of last year. That does not include construction or other companies that work there. They added a bunch this year with the new cells and drive units coming online.
They will have a minimum of 6,500 when finished but more likely 10,000.
Runarbt
Active Member
Ulmo
Active Member
About Elon factories in general:
I've seen in their Fremont "half of GF1"*, that cars are made by literally rolling (usually flying thru aerial tracks) them down assembly lines, to various stations which do various unique (to that station) things to each car. This is called the assembly "line". I've also heard that they're building a new Model 3 assembly "line", which indicates to me that indeed it will be another line in actuality, with the product going down the line being worked on again by stations.
I've often wondered how a hybrid approach with a totally different concept would work. I've posted as much before, but now I'm not just making it a side curiosity.
First, let me describe the alternative view. In my primary building block of an idea, the product being built (car, truck, spaceship, rocket part, battery, solar panel, whatever, could be generalized) would sit in one spot, and parts, materials and machines to build the product would flow into the product from disparate parts of the factory. Instead of flowing a huge multi-ton product through the factory, they can flow much lighter and often much smaller parts through multiple pathways to a product building platform. Since many types of operations would need to be performed, the robots working on the vehicle would be built to be able to do a combination of two things: (a) the robots would have a variety of generalized capabilities that are needed for building, but only those capabilities that make sense for that robot to efficiency carry. (b) the robots would have mobility, to move around the product being built, and to move to other products being built if it is in the way of other differently capable robots currently needed or otherwise not needed during some stage of building for one product, and if it was needed for other products. The heavier robots could be on tracks, and the lighter ones on wheels, and things needed could be served in the building pad areas, such as mounts to stay still, electrical and material hookups, etc.. The product would stay still, and a swarm of tools and robots (whatever words anybody cares to come up with) would work on the product, coming and going as needed.
Flowing into the end product being built would be components that themselves were also built in the factory, quite possibly on either their own pad-type build platforms or on their own miniature assembly lines, using the below described hybrid approach (keep reading). This would form a sort of tree pattern of parts flow diagram for each end product.
Now, once this view has been perfected, a hybrid approach could be entertained, where the products basically don't move much, but could have a much smaller number of "stations" as in an assembly line of current, perhaps one tenth as many stations or even 1/40th the number of stations. In this case, the different stations would be used in order to allow closer pathways for materials and parts flowing into the product without as much congestion, and would allow specialized product mover robots to occasionally come to a product and slowly move it to the next station, without having to have the super fast super heavy duty product mover that a full on assembly line ala GF1A (gigafactory 1 auto such as Fremont) has now. I would tend to say this is the deprecated way of doing things, but that it would be done only as optimal anyway in the hybrid approach. For instance, for now, paint shop would be paint shop, and they wouldn't have an environmental tent robot with environmental filter and environmental vacuum robots prepping for paint at a build pad; they'd probably just get the product moved to paint, or at least the component parts.
Of course, multiple building pads (as you should have inferred from my descriptions above) would be working in parallel, so the total output would not be limited by one "assembly line" (this should have been obvious).
That's my view. Features of the moving tools and materials assembly system (as opposed to the moving products assembly line system) include lighter movement of pieces, tighter use of 3D space (since materials and parts flows can come in from above, and so can tools), and a much more flexible product building capability without retooling; you can practically build anything of any size using this approach, if the build pads are flexible enough to be bigger at least on two connected sides (or covers on floor supplies such as some mounts, electrical plugs, etc.). This way, a 40 ton truck can be built next to a motorcycle or home battery, or in the same pad, from moment to moment, and no assembly line constraints ever come into play. If a new tool is needed for the factory, you basically provision it at the "tool input" end of the factory, and if a tool ever needs servicing, some robot fetches it (if the tool can't come on its own), and it can be serviced in a tool servicing area. Materials pour in, and products pour out.
* Elon said all future GF's will have both battery and car production per factory (pour raw Earth materials in, out comes cars (and Energy products), all-encompassing), so Nevada's GF1 is half of the GF1, the other half being Fremont factory.
Today, I played with the idea of calling Tesla's Fremont factory GF1A, and Sparks Nevada GF as GF1B, using the letters as sequences (in the A B C sequence). I first failed to see useful associations for the letters, until it became abundantly clear (I'm so slow these days): A for Automobile and B for Battery (duh!). In my parts flow factory idea rather than the product flow assembly line idea being used at GF1A, it would be better for the whole thing to be under one roof, so for instance, GF2 hopefully would just flow parts in-building without having to interstate transport parts or even any type of public infrastructure at all.
I've seen in their Fremont "half of GF1"*, that cars are made by literally rolling (usually flying thru aerial tracks) them down assembly lines, to various stations which do various unique (to that station) things to each car. This is called the assembly "line". I've also heard that they're building a new Model 3 assembly "line", which indicates to me that indeed it will be another line in actuality, with the product going down the line being worked on again by stations.
I've often wondered how a hybrid approach with a totally different concept would work. I've posted as much before, but now I'm not just making it a side curiosity.
First, let me describe the alternative view. In my primary building block of an idea, the product being built (car, truck, spaceship, rocket part, battery, solar panel, whatever, could be generalized) would sit in one spot, and parts, materials and machines to build the product would flow into the product from disparate parts of the factory. Instead of flowing a huge multi-ton product through the factory, they can flow much lighter and often much smaller parts through multiple pathways to a product building platform. Since many types of operations would need to be performed, the robots working on the vehicle would be built to be able to do a combination of two things: (a) the robots would have a variety of generalized capabilities that are needed for building, but only those capabilities that make sense for that robot to efficiency carry. (b) the robots would have mobility, to move around the product being built, and to move to other products being built if it is in the way of other differently capable robots currently needed or otherwise not needed during some stage of building for one product, and if it was needed for other products. The heavier robots could be on tracks, and the lighter ones on wheels, and things needed could be served in the building pad areas, such as mounts to stay still, electrical and material hookups, etc.. The product would stay still, and a swarm of tools and robots (whatever words anybody cares to come up with) would work on the product, coming and going as needed.
Flowing into the end product being built would be components that themselves were also built in the factory, quite possibly on either their own pad-type build platforms or on their own miniature assembly lines, using the below described hybrid approach (keep reading). This would form a sort of tree pattern of parts flow diagram for each end product.
Now, once this view has been perfected, a hybrid approach could be entertained, where the products basically don't move much, but could have a much smaller number of "stations" as in an assembly line of current, perhaps one tenth as many stations or even 1/40th the number of stations. In this case, the different stations would be used in order to allow closer pathways for materials and parts flowing into the product without as much congestion, and would allow specialized product mover robots to occasionally come to a product and slowly move it to the next station, without having to have the super fast super heavy duty product mover that a full on assembly line ala GF1A (gigafactory 1 auto such as Fremont) has now. I would tend to say this is the deprecated way of doing things, but that it would be done only as optimal anyway in the hybrid approach. For instance, for now, paint shop would be paint shop, and they wouldn't have an environmental tent robot with environmental filter and environmental vacuum robots prepping for paint at a build pad; they'd probably just get the product moved to paint, or at least the component parts.
Of course, multiple building pads (as you should have inferred from my descriptions above) would be working in parallel, so the total output would not be limited by one "assembly line" (this should have been obvious).
That's my view. Features of the moving tools and materials assembly system (as opposed to the moving products assembly line system) include lighter movement of pieces, tighter use of 3D space (since materials and parts flows can come in from above, and so can tools), and a much more flexible product building capability without retooling; you can practically build anything of any size using this approach, if the build pads are flexible enough to be bigger at least on two connected sides (or covers on floor supplies such as some mounts, electrical plugs, etc.). This way, a 40 ton truck can be built next to a motorcycle or home battery, or in the same pad, from moment to moment, and no assembly line constraints ever come into play. If a new tool is needed for the factory, you basically provision it at the "tool input" end of the factory, and if a tool ever needs servicing, some robot fetches it (if the tool can't come on its own), and it can be serviced in a tool servicing area. Materials pour in, and products pour out.
* Elon said all future GF's will have both battery and car production per factory (pour raw Earth materials in, out comes cars (and Energy products), all-encompassing), so Nevada's GF1 is half of the GF1, the other half being Fremont factory.
Today, I played with the idea of calling Tesla's Fremont factory GF1A, and Sparks Nevada GF as GF1B, using the letters as sequences (in the A B C sequence). I first failed to see useful associations for the letters, until it became abundantly clear (I'm so slow these days): A for Automobile and B for Battery (duh!). In my parts flow factory idea rather than the product flow assembly line idea being used at GF1A, it would be better for the whole thing to be under one roof, so for instance, GF2 hopefully would just flow parts in-building without having to interstate transport parts or even any type of public infrastructure at all.
Last edited:
AZ Desert Driver
Rare combination
This is kinda like how I run my home garage on small projects ...Just a workbench where I mount the item in a vice, and bring in parts and tools to do various stages of the project. Continuing the "workbench" idea, I can not do all the things at one station - I often have to move the (project) from bench to floor for more room. I cant see painting a car in the same place welders stuck the frame together.
I applaud your creative thinking, but it seems like a throw back to before 1920's Fords assembly line where "workbench" was the model to replace.
Similar threads
- Article
