Climbs most stairs which occur in daily life:
35 degrees of elevation, each step height up to 18 cm
Carries heavy loads easily:
- Payload up to 6 kg
- Can climb a 5-story or higher building
Gives a sense of familiarity and safety:
- Flexible body materials with a soft touch
- Shock absorbing body, safe for humans and furniture
"Stairs" for Humans can be
"Unexplored Frontiers" for Robots
Our houses are full of stairs with variable dimensions and courses.
However it is impossible for robots to know these parameters until they have actually climbed them.
In order to climb new stair environments in one-shot, robots had to have a very high costed feedback system consisting of :
The "Softness" of Amoebas,
now integrated in Robotics
An Amoeba can adapt to versatile environments through its soft body.
Inspired by the Amoeba, we adapted elastic material to a crawler.
The deformation of this soft crawler acts as an integrated system of sensors, processors, and actuators:
In this way, soft crawlers can achieve the same functionality as feedback control
with much lower computational and financial costs.
This is the way how the Amoeba survives in nature, and we have adapted this idea to create our soft crawlers.
AMOEBA ENERGY's CEO, Dr. Masashi Aono, also an associate professor at Keio University,
has been studying the sophisticated information processing capacities of
a single-celled amoeboid organism (the true slime mold Physarum).
He has shown that this Amoeba is able to find solutions to complex optimization problems, through changing its shape in response to environmental stimuli.
The "Traveling salesman problem" is one such example, known to be increasingly important when promoting the efficiency of various real-world systems.
The aim is to create new types of problem solvers (non-Neumann computers), implementing the algorithms inspired by the AMOEBA's adaptive shape-changing dynamics.
Our team of researchers from institutes including Keio University, Hokkaido University, Tokyo Institute of Technology,
and Yokohama National University, is developing a compact and low-power electronic circuit called the "Amoeba-inspired Computing Chip,"
which mimics the Amoeba's information processing dynamics.
Exploiting the physical process of parallel electric current, it can quickly find solutions to complex problems while avoiding the "combinatorial explosion" problem.
When incorporated in the robot as its brain, this chip has the potential to produce "creativity" to discover how to overcome unknown obstacles without any pre-programmed instructions.