From ASIMO to Pepper to Sophia: A Short History of Humanoid Robots Before This Boom

The current wave didn't come from nowhere. Three very different robots shaped how the public thinks about humanoids long before Optimus existed.

Tesla, Figure, and Unitree get most of today's attention, but the public's mental image of a "humanoid robot" was shaped by machines that predate this boom by a decade or more — and their very different fates are worth understanding.

ASIMO: technical triumph, commercial dead end

Honda's ASIMO ran, climbed stairs, and poured drinks in public demos starting in 2000, becoming the default image of a humanoid robot for most of two decades. Honda retired its development program in 2018 without ever turning it into a sustained commercial product — extraordinary engineering that arrived before a viable business model existed to support it.

Pepper: real deployment, real limits

SoftBank's Pepper — technically a wheeled upper-body humanoid rather than a full biped — was genuinely deployed at scale: thousands of units in retail stores, banks, and airports worldwide for greeting and simple customer interaction. Production ended in 2021 after the robot struggled to prove lasting commercial value beyond initial novelty, a real cautionary data point about the gap between deployment numbers and sustained business value.

Sophia: fame without a factory

Hanson Robotics' Sophia became a global media phenomenon after being granted Saudi Arabian citizenship in 2017 — a publicity milestone that did more than any technical achievement to put humanoid robots into mainstream conversation. Sophia is built primarily for expressive, conversational interaction rather than mobility or manipulation, closer in spirit to Engineered Arts' Ameca than to an industrial worker robot.

Why this history matters now

All three demonstrate the same underlying lesson from different angles: technical achievement, real deployment scale, and public fame are three separate things, and a robot can win big on one axis while falling short on the others. The current generation of humanoid companies is, in large part, trying to be the first to win on all three simultaneously.