Artificial Intelligence – Robotics
Part 1 – What is robotics?
Some examples:
Vaucanson’s Canard Digérateur: 1739; the Digesting Duck; a Robotic Duck:
This duck could quack, drink water, eat grain, digest it, and then poo it out.
It contained a ‘chemical laboratory’.
A mechanical duck that could eat was special because it replicated some aspects of
an organism.
People speculated how it worked and what was going on inside the duck, but what
really was going on was much different.
For a very long time, people have been trying to mechanise aspects of nature, AI is
just taking this idea further, but the idea is therefore not new.
W. Grey Walter’s autonomous robots: 1948
He was interested in the ‘mimicry of life’; using basic materials, such as cogs from
gas meters, he constructed a series of mobile robots that resembled tortoises.
He for example used a shell as bump attack.
One of his famous classic experiments was Elsie;
- Elsie was attracted to light, but avoided very bright light.
- Elsie would move away from the recharging station until her battery ran low,
then return, and recharge herself.
- Walter wondered if the behaviour of this robots ‘might be accepted as
evidence of some degree of self-awareness
Elsie had a photo preceptor with which she was attracted to light.
(Can it be called self-awareness that Elsie is able to recharge itself, so able to
recognise that it is empty, and then refill itself.
This is a very early example of a mobile autonomous robot.
SRI’s Shakey: 1966
- Shakey inhabited blocks world; a sanitised environment, a restrained area of
the real world. A version of the real world
- Shakey was a landmark (important turning point): with integrated planning
and reasoning with robotics, until then, robotics could only perform specific
tasks
- Shakey could solve tasks that involved moving blocks around the environment
- Problem; Shakey could become confused due to discrepancies between his
model of the world and what he finds in the actual world → therefore this
integrating of different areas could go wrong
- Shakey suffered from a lack of robustness (ability to withstand or overcome
adverse conditions or rigorous testing)
Putting everything that people knew about AI in one robot was new
It was a huge/important attempt to integrate the different areas of AI, and the goal
was to solve problems by moving around these blocks.
Part 1 – What is robotics?
Some examples:
Vaucanson’s Canard Digérateur: 1739; the Digesting Duck; a Robotic Duck:
This duck could quack, drink water, eat grain, digest it, and then poo it out.
It contained a ‘chemical laboratory’.
A mechanical duck that could eat was special because it replicated some aspects of
an organism.
People speculated how it worked and what was going on inside the duck, but what
really was going on was much different.
For a very long time, people have been trying to mechanise aspects of nature, AI is
just taking this idea further, but the idea is therefore not new.
W. Grey Walter’s autonomous robots: 1948
He was interested in the ‘mimicry of life’; using basic materials, such as cogs from
gas meters, he constructed a series of mobile robots that resembled tortoises.
He for example used a shell as bump attack.
One of his famous classic experiments was Elsie;
- Elsie was attracted to light, but avoided very bright light.
- Elsie would move away from the recharging station until her battery ran low,
then return, and recharge herself.
- Walter wondered if the behaviour of this robots ‘might be accepted as
evidence of some degree of self-awareness
Elsie had a photo preceptor with which she was attracted to light.
(Can it be called self-awareness that Elsie is able to recharge itself, so able to
recognise that it is empty, and then refill itself.
This is a very early example of a mobile autonomous robot.
SRI’s Shakey: 1966
- Shakey inhabited blocks world; a sanitised environment, a restrained area of
the real world. A version of the real world
- Shakey was a landmark (important turning point): with integrated planning
and reasoning with robotics, until then, robotics could only perform specific
tasks
- Shakey could solve tasks that involved moving blocks around the environment
- Problem; Shakey could become confused due to discrepancies between his
model of the world and what he finds in the actual world → therefore this
integrating of different areas could go wrong
- Shakey suffered from a lack of robustness (ability to withstand or overcome
adverse conditions or rigorous testing)
Putting everything that people knew about AI in one robot was new
It was a huge/important attempt to integrate the different areas of AI, and the goal
was to solve problems by moving around these blocks.
