Belts vs. Bots UPS Efficiency in Space Age – What's the Current Paradigm?

[MHplays]

I've always been an avid belt user, mainly because I’ve had an intuitive sense that belts are more UPS-efficient than having a swarm of bots flying around the map. Plus, with my PC being on the weaker side, I’ve leaned heavily towards belts, with about 90% of my base relying on them.

Recently, I came across the belts vs. bots UPS efficiency debate, and to my surprise, there were some compelling arguments suggesting that bots might actually be more UPS-efficient than belts. However, most of these discussions happened long before the Space Age update. This got me thinking—what’s the current paradigm in the belts vs. bots UPS efficiency debate, especially after all the recent changes?

For Belts in Space Age they got:

• Turbo Belts, with higher items per second throughput.
• Stack Inserter, wich Greatly boosts the throughput of belts.
• Quality Inserters, with Faster rotation speed and quicker item insertion.

And for Bots they got:

• Improved Navigation and Recharging Algorithms, likely enhance overall performance.
• Quality Robots, with better batteries, longer flight reach, and less need to recharge, optimizing their efficiency.
• Quality Roboports, with faster robot recharging, further optimizing bot performance.

Other Factors Impacting Both:

• Quality Chests: Higher capacity.
• Elevated Rails: Improve train logistics.
• Quality Fuels: Higher max speed and acceleration for trains.

Remind me if I missed any other significant factors that could influence this debate.

I’m curious to hear your thoughts and experiences with belts and bots in the Space Age. Which of them do you think received the most significant efficiency upgrades? What’s the community’s current stance on this debate?

[waterBear]

Without running the numbers, it's hard to believe that bots are more UPS efficient. We are late-game in our playthrough so we have nothing but (many thousands of) legendary bots and roboports, and their throughput is still much, much lower than, say, a few stacked turbo belts. My intuition is that throughput-per-update is much higher for belts. Even if you need 3x the updates for a belt, if it carries 15x the items per second, it still wins.

Bots also have all the same limitations they've always had. Large and / or convex networks are still bad, bot clustering leading to long delays, etc. They are extremely useful in certain situations, but not in every situation.

[Khazul]

I find that when you have effectively a single huge logistics network, that may be true, but when it comes to isolated logistics networks (ie the orange areas dont touch) with sufficient ports and bots to ensure there is always sufficient immediately available/active, bot throughput can be very high easily able to keep up with legendary production machines and inserters without hammering UPS. With large network covering an entire base, then I find they can get too easily distracted, or end up idling in the wrong place etc with the consequence that throughput often gets hammered somewhere.

I tend to use a mix of belts and bots and trains for these reasons. My isolated zones with bots are mostly used for the highest per-machine throughput scenarios.