Past And The Present: The History And Evolution Of Robots

The idea of a self-acting machine goes back to Archytas of Tarentum, who built a steam-powered wooden pigeon around 350 BC. The modern story starts in 1961, when George Devol and Joseph Engelberger installed Unimate, the first industrial robot, at a New Jersey GM plant. Robots since then have moved from factory arms to Mars rovers, social robots like Sony AIBO and Nadine, and today's humanoid robots such as Tesla Optimus and Figure 02.

The word ‘robot’ conjures up an array of images for most people, from R2D2 and C3PO of Star Wars fame, to the hyper-masculine Arnold Schwarzenegger in the Terminator series or the Rover Sojourner, which explored the Martian landscape as part of the Mars Pathfinder mission.

The idea of self-acting machines is, surprisingly, almost as old as recorded science. Around 350 BC, the Greek philosopher and engineer Archytas of Tarentum is said to have built a wooden pigeon propelled by jets of steam or compressed air, often cited as the first “robot” in history. Centuries of clockwork automata followed, but the modern story really begins in 1936, when Alan Turing famously introduced the concept of a theoretical computer, the Turing Machine.

With the arrival of programmable computers in the 1940s, the first real robots began to take shape. The earliest robots as we know them today were the work of George Devol, who in 1954 filed a patent for a reprogrammable manipulator (basically a robot) called Unimate; the patent was granted in 1961. In 1956, Devol met engineer Joseph Engelberger at a cocktail party, and the two co-founded Unimation, the world’s first robotics company, to commercialize the design. The first Unimate arm was installed in 1961 on the General Motors die-cast line at Inland Fisher Guide in Ewing Township, New Jersey, where it unloaded hot, dangerous parts that human workers had previously had to handle. For his role in turning the patent into a real industry, Engelberger is now known as “the Father of Robotics”.

In 1966, the Stanford Research Institute created Shakey, the first general-purpose mobile robot able to reason concerning its own actions. While other robots would have to be instructed on each individual step in order to complete a larger task, Shakey could analyze commands and break them down into basic chunks by itself. These actions involved traveling from one location to another, turning light switches on and off, opening and closing the doors, climbing up and down from rigid objects, and pushing movable objects around.

Three years later, Victor Scheinman, a Mechanical Engineering student working in the Stanford Artificial Intelligence Lab (SAIL) created the Stanford Arm. The arm’s design became a standard for the design of future robot arms. In fact, it is still being used today.

In 1974, MIT engineer David Silver built the Silver Arm, capable of assembling small parts together using touch sensors, which pushed robotics into more delicate, dexterous territory.

In 1985, a robotic arm called the PUMA 200 was used by Dr. Yik San Kwoh in Long Beach, California, to orient a needle for a CT-guided brain biopsy. It was the first time a robot was used in human surgery. The PUMA 200 was later retired from the operating room over safety concerns, but it opened the door to modern surgical robots.

In the early 1990s, Stanford neurosurgeon Dr. John Adler designed the CyberKnife, a robotic radiosurgery system that uses X-rays to locate tumors and then targets them with focused beams of radiation. He founded Accuray in 1992 to commercialize it, and the first patient was treated at Stanford in 1994.

1997 was perhaps one of the most important years for robotics. It was the year of the Mars Pathfinder mission. The robotic Mars rover Sojourner was released onto Mars’ surface and its job was to transmit data back to Earth from the Martian soil.

In the year 1999, Sony released AIBO, a robotic dog with the ability to learn, entertain and communicate with its owner.

With further advancements in technology, robots started becoming more and more autonomous. In 2002, Sony unveiled the SDR-4X, a small humanoid robot that could remember faces, dance, and sing in harmony with other SDR-4X units (it later evolved into the QRIO).

Soon, inventors started making robots more human-like. In recent years, features like artificial skin tissues, emotions, self-awareness and autonomous thinking have been incorporated in robots. Robothespian is one such machine that is capable of entertaining and communicating with people. It is fully interactive and multilingual.

In 2015, Nanyang Technological University Singapore unveiled Nadine, at the time one of the most lifelike social robots ever built (and modeled on her creator, Prof. Nadia Magnenat Thalmann). Nadine operated as a university receptionist, and scientists continue to work on robots that can provide childcare and offer companionship to lonely elderly people.

Perhaps one of the most life-changing applications of robotics is the wearable suit known as HAL (Hybrid Assistive Limb), developed by Prof. Yoshiyuki Sankai at the University of Tsukuba and sold by Cyberdyne Inc. HAL doesn't actually read the brain directly; instead, electrodes on the skin pick up the faint bio-electric signals the brain sends to the muscles, and the suit's motors move the limbs in step with the wearer's intent. That makes it especially useful for people with mobility problems and the elderly.

The decade after 2015 turned out to be the busiest in robotics history. In June 2020, Boston Dynamics began selling its four-legged Spot robot to the public, the first commercially available robot from the company. On 19 April 2021, NASA’s tiny Ingenuity helicopter, dropped off by the Perseverance rover, performed the first powered flight on another planet, hovering for 39 seconds in the thin Martian atmosphere. In 2022, Tesla unveiled its Optimus humanoid prototype, kicking off a race to build general-purpose human-shaped robots for factories and homes. By 2024, Boston Dynamics had retired its hydraulic Atlas and replaced it with a sleek all-electric version, Figure AI was deploying its Figure 02 humanoids at a BMW plant in Spartanburg, and China’s Unitree released the H1 and G1 humanoids at prices that started under $20,000. In parallel, a 2023 model from Google DeepMind called RT-2 showed that large language models could let robots translate web-scale knowledge into real-world actions, accelerating the shift from carefully scripted machines to robots that can reason their way through new tasks.

Robots are doing more good than harm in terms of improving our quality of life. A few years down the line, just like mobile phones, robots might become an integral part of our lives. The question is… is that what we really want?