Jay Smith @jaynemesis: "𝗥𝗼𝗯𝗼𝘁 𝗪𝗮𝗿𝘀: 𝗔𝗿𝗲 𝗛𝘂𝗺𝗮𝗻𝗼𝗶𝗱 𝗥𝗼𝗯𝗼𝘁𝘀..."

United Kingdom

𝗥𝗼𝗯𝗼𝘁 𝗪𝗮𝗿𝘀: 𝗔𝗿𝗲 𝗛𝘂𝗺𝗮𝗻𝗼𝗶𝗱 𝗥𝗼𝗯𝗼𝘁𝘀 𝗥𝗲𝗮𝗹𝗹𝘆 𝗨𝘀𝗲𝗳𝘂𝗹? $ABBN.ZU (ABB LTD) $NVDA (NVIDIA Corporation) $TSLA (Tesla Motors, Inc.) $SIE.DE (Siemens Aktiengesellschaft) $1810.HK (Xiaomi Corp) //𝘗𝘢𝘳𝘵 𝟣 𝘰𝘧 𝟤 - 𝘚𝘦𝘦 𝘤𝘰𝘮𝘮𝘦𝘯𝘵𝘴 𝘧𝘰𝘳 𝘱𝘢𝘳𝘵 𝟤 Robots are cool - think $TSLA Optimus dancing and Boston Dynamics’ Spot on a factory tour. But beyond the fun demos, are the humanoid robots promising to revolutionize our workplaces and homes by emptying the trash and stacking supermarket shelves really the solution? Humanoid Robots are like a Swiss Army Knife, impressive engineering, but who’s the buyer? 𝙄𝙣𝙙𝙪𝙨𝙩𝙧𝙞𝙖𝙡 𝙃𝙮𝙥𝙚 𝙫𝙨 𝙍𝙚𝙖𝙡𝙞𝙩𝙮 Humanoids are marketed as the perfect alternative to humans in warehouses and factories globally. No breaks or time off, lower cost, higher uptime and multi-taskers once trained. 𝐻𝑒𝑟𝑒’𝑠 𝑡ℎ𝑒 𝑟𝑒𝑎𝑙𝑖𝑡𝑦 𝑐ℎ𝑒𝑐𝑘: Most major industrial facilities are now well standardized and structured for efficiency, they have repeatable workflows that are more suited to specialized robots than humanoids. Why would companies invest in robots that have so many additional parts for fingers and thumbs, legs, knees and feet, when they could just use gripping jaws and wheels? Then there’s the capital expenditure and expertise required to shift from human driven operations, to humanoids with teams of engineers to keep it all running. Most tasks that can be automated already have been, and the ones that haven’t rely upon human judgement and adaptability. 𝙃𝙪𝙢𝙖𝙣𝙤𝙞𝙙𝙨 𝙫𝙨 𝙃𝙪𝙢𝙖𝙣𝙨: 𝘾𝙖𝙣 𝙏𝙝𝙚𝙮 𝙍𝙚𝙖𝙡𝙡𝙮 𝙍𝙚𝙥𝙡𝙖𝙘𝙚 𝙐𝙨? Let’s explore how well humanoid robots stack up against what they are trying to imitate. 𝐷𝑜𝑤𝑛𝑡𝑖𝑚𝑒 & 𝐵𝑟𝑒𝑎𝑘𝑠: Although robots don’t need sick days, they do still have downtime. They require time for software updates, hardware inspection, cleaning, maintenance & calibration, repairs, and charging. Humans typically work 8 hours per day, 5 days a week. Including sick leave, public holidays and paid time off, it’s around 250 days per year. Humanoids, once set up, can work up to 22 hours per day 6 days a week, with 2 hours daily for charging and inspections and a full day for software updates, cleaning, repairs, etc. Humanoids win on uptime. 𝐶𝑜𝑠𝑡: Humans in emerging markets like Vietnam, China, Indonesia, etc typically get paid around $3/hr. They don’t consume any electricity, there are no capital costs to be amortized, they don’t require spare parts and the support infrastructure (HR, break rooms, Management Staff, Health & Safety, Administration work, etc) comes to around a 12.5% overhead or $0.40/hr. Salary: $3.00/hr Support/Infrastructure: $0.40/hr Total: $3.40/hr Humanoids don’t get a salary, consume around $0.21/hr of electricity, need approximately $0.14/hr for spare parts, and require support infrastructure (Maintenance Bays, Software & Hardware Technicians, Management Platform, AI & Compute, Analysts & Administrators, etc) at $0.26/hr. The Humanoids could cost as little as $30,000 in 2030, meaning an Amortization cost of $0.63/hr. Energy: $0.21/hr Spare Parts: $0.14/hr Support/Infrastructure: $0.26/hr Capex: $0.63/hr Total: $1.24/hr If the assumption that Humanoids are just $30,000 each is correct, humanoids comfortably win on cost. However, if for example the cost only falls to $60,000 by 2030 the total cost would be $1.87/hour, which isn’t quite as compelling. This also doesn’t account for the cost of capital. The funds used to purchase the robots would likely utilize some sort of loan, so the true cost for humanoids will be higher in most cases. 𝐸𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑐𝑦 & 𝐸𝑓𝑓𝑒𝑐𝑡𝑖𝑣𝑒𝑛𝑒𝑠𝑠: While the above metrics are easy to measure, this one is much more difficult and varies dramatically based on task and working environment. It can easily be argued that until AI improves substantially, humans are much better at handling edge cases. Irregularly shaped objects, clutter, spillages, etc. Alongside this, new tasks could take days or even weeks to train humanoid robots to carry out effectively, requiring simulations and testing. Some more intricate tasks are better suited to humans, we can fit in small spaces and have better dexterity and coordination than humanoids do currently. Finally, there’s accidents and safety. Humans are better at spotting and avoiding hazards, and can call emergency stops when appropriate. Humanoids, either require heavy monitoring for this, or could suffer from false positives, slowing down production, or not recognizing risks, resulting in collisions or costly errors. Based on the current technology, and where I expect it will be in 2030, Humans win this category. 𝑆𝑢𝑚𝑚𝑎𝑟𝑦: It appears inevitable that eventually, Humanoids will surpass Humans in all but edge cases. However, self-driving cars serve as a perfect example of a technology that has actually taken much longer to mature than most investors and leaders anticipated. //𝘌𝘯𝘥 𝘰𝘧 𝘱𝘢𝘳𝘵 𝟣 𝘰𝘧 𝟤 - 𝘚𝘦𝘦 𝘤𝘰𝘮𝘮𝘦𝘯𝘵𝘴 𝘧𝘰𝘳 𝘱𝘢𝘳𝘵 𝟤