Page 1 of 1

Logistic Numbers

Posted: Tue May 01, 2018 1:19 am
by Boldar
Hi,
a while ago i found a calculation about the ideal ratio of robots per roboport, calculated dependent of Robot Speed Tech. For some reason, i cannot find it again.
Has anyone correct numbers for 0.16, assuming continious robot usage, or the formular to calculate it from robot speed and charging speed?

Re: Logistic Numbers

Posted: Tue May 01, 2018 2:01 am
by bobucles
There isn't really an ideal ratio. If you have around 50(high lvl bots) to 100(low lvl bots) busy bots per roboport you can be reasonably safe against long recharge lines. But the amount of charging you need is entirely dependent on the amount of bot traffic going through an area. If there are bots piling up in rings around your roboports, that's your cue to build more.

Re: Logistic Numbers

Posted: Tue May 01, 2018 2:24 am
by Boldar
No, there is an ideal ratio - that is the ratio where on a long-term view enough charge-pads will be available. For calculating it you need to look at the bot-duty-cycle. E.g. if each bot charges 10 seconds and then flies 10 seconds, each roboport could support 8 bots (4 charging slots / 0.5 duty cycle). The only problem is to get the duty cycle, because you need the charge-speed for that as well as the bots power consumption per time unit.
I was just hoping someone already had the numbers.


Edit: i did some calculations and came to this (please correct me if you see mistakes):

Code: Select all

Tech	robot speed	draw	Battery	work time	Charge time	Duty cycle	Bots per Roboport
							
 5		10,20		 54,00		1500,00		27,78	1,50		18,52			74,07
 6		12,15		 63,75		1500,00		23,53	1,50		15,69			62,75
 7		14,10		 73,50		1500,00		20,41	1,50		13,61			54,42
 8		16,05		 83,25		1500,00		18,02	1,50		12,01			48,05
 9		18,00		 93,00		1500,00		16,13	1,50		10,75			43,01
10		19,95		102,75		1500,00		14,60	1,50		9,73			38,93
11		21,90		112,50		1500,00		13,33	1,50		8,89			35,56
12		23,85		122,25		1500,00		12,27	1,50		8,18			32,72
13		25,80		132,00		1500,00		11,36	1,50		7,58			30,30
14		27,75		141,75		1500,00		10,58	1,50		7,05			28,22
15		29,70		151,50		1500,00		 9,90	1,50		6,60			26,40
16		31,65		161,25		1500,00		 9,30	1,50		6,20			24,81
17		33,60		171,00		1500,00		 8,77	1,50		5,85			23,39
18		35,55		180,75		1500,00		 8,30	1,50		5,53			22,13
19		37,50		190,50		1500,00		 7,87	1,50		5,25			21,00
20		39,45		200,25		1500,00		 7,49	1,50		4,99			19,98
21		41,40		210,00		1500,00		 7,14	1,50		4,76			19,05
22		43,35		219,75		1500,00		 6,83	1,50		4,55			18,20
23		45,30		229,50		1500,00		 6,54	1,50		4,36			17,43
24		47,25		239,25		1500,00		 6,27	1,50		4,18			16,72
25		49,20		249,00		1500,00		 6,02	1,50		4,02			16,06
26		51,15		258,75		1500,00		 5,80	1,50		3,86			15,46
27		53,10		268,50		1500,00		 5,59	1,50		3,72			14,90
28		55,05		278,25		1500,00		 5,39	1,50		3,59			14,38
29		57,00		288,00		1500,00		 5,21	1,50		3,47			13,89
30		58,95		297,75		1500,00		 5,04	1,50		3,36			13,43
...
100		195,45		980,25		1500,00		1,53	1,50		1,02			 4,08

So as you can see, for robot working speed 100 you need one roboport for every 4 robots.

Re: Logistic Numbers

Posted: Tue May 01, 2018 3:42 am
by Zavian
Well as you unlock faster bots, you need less of them to do the same amount of work, but people typically don't remove bots from already working outposts. So the figure I've heard before is one roboport can support about 40 bots. That probably assumes you have a layout with good pathing, and that on average bots will be evenly spread over all roboports. If you have a large layout with many bots wanting to charge in one area then that area will need most of the roboports.

Re: Logistic Numbers

Posted: Tue May 01, 2018 8:34 pm
by Zanthra
Boldar wrote:No, there is an ideal ratio - that is the ratio where on a long-term view enough charge-pads will be available. For calculating it you need to look at the bot-duty-cycle. E.g. if each bot charges 10 seconds and then flies 10 seconds, each roboport could support 8 bots (4 charging slots / 0.5 duty cycle). The only problem is to get the duty cycle, because you need the charge-speed for that as well as the bots power consumption per time unit.
I was just hoping someone already had the numbers.


Edit: i did some calculations and came to this (please correct me if you see mistakes):

Code: Select all

Tech	robot speed	draw	Battery	work time	Charge time	Duty cycle	Bots per Roboport
							
 5		10,20		 54,00		1500,00		27,78	1,50		18,52			74,07
 6		12,15		 63,75		1500,00		23,53	1,50		15,69			62,75
 7		14,10		 73,50		1500,00		20,41	1,50		13,61			54,42
 8		16,05		 83,25		1500,00		18,02	1,50		12,01			48,05
 9		18,00		 93,00		1500,00		16,13	1,50		10,75			43,01
10		19,95		102,75		1500,00		14,60	1,50		9,73			38,93
11		21,90		112,50		1500,00		13,33	1,50		8,89			35,56
12		23,85		122,25		1500,00		12,27	1,50		8,18			32,72
13		25,80		132,00		1500,00		11,36	1,50		7,58			30,30
14		27,75		141,75		1500,00		10,58	1,50		7,05			28,22
15		29,70		151,50		1500,00		 9,90	1,50		6,60			26,40
16		31,65		161,25		1500,00		 9,30	1,50		6,20			24,81
17		33,60		171,00		1500,00		 8,77	1,50		5,85			23,39
18		35,55		180,75		1500,00		 8,30	1,50		5,53			22,13
19		37,50		190,50		1500,00		 7,87	1,50		5,25			21,00
20		39,45		200,25		1500,00		 7,49	1,50		4,99			19,98
21		41,40		210,00		1500,00		 7,14	1,50		4,76			19,05
22		43,35		219,75		1500,00		 6,83	1,50		4,55			18,20
23		45,30		229,50		1500,00		 6,54	1,50		4,36			17,43
24		47,25		239,25		1500,00		 6,27	1,50		4,18			16,72
25		49,20		249,00		1500,00		 6,02	1,50		4,02			16,06
26		51,15		258,75		1500,00		 5,80	1,50		3,86			15,46
27		53,10		268,50		1500,00		 5,59	1,50		3,72			14,90
28		55,05		278,25		1500,00		 5,39	1,50		3,59			14,38
29		57,00		288,00		1500,00		 5,21	1,50		3,47			13,89
30		58,95		297,75		1500,00		 5,04	1,50		3,36			13,43
...
100		195,45		980,25		1500,00		1,53	1,50		1,02			 4,08

So as you can see, for robot working speed 100 you need one roboport for every 4 robots.

I read that as supporting 12 robots per roboport at robot speed 100. If they work for 1.53 seconds, then charge for 1.5 seconds, a single robot charging point would only charge that robot for 1.5 seconds out of every 3.03 or about 49% of the time. With 2 robpots per charge point, each charge point goes up to 98% of the time spent charging a robot, or 3 seconds out of 3.03, but only once you get to 3 robots per charge point does it goes to 100%. 3 robots per charge point, 4 points per roboport, makes 12 robots per port.

I do feel that this model is only applicable to a very limited number of situations, where all your roboports, source chests, and destination chests are very close to each other.

Re: Logistic Numbers

Posted: Tue May 01, 2018 9:15 pm
by bobucles
I do feel that this model is only applicable to a very limited number of situations, where all your roboports, source chests, and destination chests are very close to each other.
That's because it assumes that your roboport network is a frictionless singularity in a perfect vacuum. It's not a terribly useful metric and it does a novice more harm than good.

Re: Logistic Numbers

Posted: Tue May 01, 2018 11:14 pm
by Boldar
But it is perfectly applicable to my train-unloading station were if have short distances and the robots are busy quitew often due to 8 16-wagon unloading spots serving 32 trains in total and unloading iron, copper, coal and stone. I can upload some pictures later.

Re: Logistic Numbers

Posted: Tue May 01, 2018 11:41 pm
by quyxkh
at robot speed 100
w. t. f.?

... oh. you mean 100km/h, not the robot speed 100 research, duh.

Still, spending 256 launches to get a <10% speed boost (92.9 to 99.9 km/h) isn't going to change any designs or save any energy, the recharge cost depends on route length not speed, right? With faster bots you can wind up needing _more_ ports. What kind of design corner can you get in to that spending that much for an incremental boost like that is the right way out?

b.t.w. advocating for robot speed 100 research, total cost ±2^94 space science, seems . . . slightly out of reach.

Re: Logistic Numbers

Posted: Tue May 01, 2018 11:58 pm
by Boldar
The example calculations i did did include robot speed research 100, at which point robots will spend 1,53s in the air and then charge for almost the same time.
But it was just as an example.