Collective behavior that emerges from the interactions between agents and with the environment are one of the most prominent examples of self-organizing systems. The field of swarm robotics allows to research and engineer such collective behavior. However to do this, there is the need for small cost-effective swarmbots.
The table below lists a number of robots used in real experiments in the field of swarm robotics. It shows the most important parameters and information about the robots simplifying comparison between them. For more detailed description there are links (right after the robots' names) to websites or articles exhibiting specifications or conducted experiments.
One of the first questions, coming up in the beginning of a project concerning swarm robotics, is money. Conducting real experiments with dozens of robots require high investments to buy all needed hardware and software. The price of the robots varies greatly that depends on a number of sensors and actuators, quality of a robot's frame, and demand for a particular robot. Unfortunately, the price of some robots is not specified. The most of these robots have an open-source design and code so they can be built from scratch in case there is such a need.
The table shows that the majority of the robots are differential drive robots whose movement is based on two wheels placed on either side of the robot. It allows achieving higher speed and precisely controlling the direction of movement. However, there are also some disadvantages, such as a need for a prepared surface without pits and bumps. Robots using the tracks usually can move on the unprepared surface, which facilitates a preparation for experiments and demonstrations. A unique type of locomotion is vibration (check the Kilobot robot) which tends to be slow and imprecise way of movement. It also requires a special smooth surface.
Size matters in the field of swarm robotics. A swarm consists of a large number of individuals which are quite effective in performing of important tasks for surviving of the colony. A significant number of robots is important in order to reproduce the swarm behavior or to check a hypothesis.
Unlike individuals in the real swarms, robots cannot generate energy from organics and still require a source of energy such as a battery. The working time of a robot depends on the capacity of battery and the amount of sensor and actuators using by a robot in experiments. Some of the robots have charging stations (e.g. Elisa, Jasmine) which may be useful in conducting of long-term experiments. The marXbot goes even further supporting exchange of the battery in less than 10 seconds without shutting down the power.
Communication in swarm robotics can play a crucial role while the effectiveness of the whole colony depends on how well the robots can send and receive messages. There are two types of communication: abstract (e.g. Wi-Fi, Bluetooth, ZigBee, RF) and situated (e.g. IR). Both of them work well for messages exchange, but situated communication provides additional information about the sender and the message, e.g. strength of the signal and its direction.
The choice of the robot depends on experiments where it will be used. Select the most important features of the future research and pick a robot that fits best.
The table below lists a number of robots used in real experiments in the field of swarm robotics. It shows the most important parameters and information about the robots simplifying comparison between them. For more detailed description there are links (right after the robots' names) to websites or articles exhibiting specifications or conducted experiments.
| Robot | Cost | Locomotion | Speed (cm/s) | Size (cm) | Battery life (hours) | Communication |
|---|---|---|---|---|---|---|
| Alice[1] | N/A | wheels | 4 | 2.1x2.1x2.1 | 10 | RF |
| Jasmine[1,2] | $118 | wheels | N/A | 2.6x2.6x2.6 | 1-2 | RF, IR |
| Elisa[1] | $390 | wheels | 60 | ∅5 | 3 | RF,IR |
| Colias[1,2] | $37 | wheels | 35 | ∅4 | 1-3 | IR |
| Kilobot[1,2] | $111 | vibration | 1 | ∅3.3 | 3-24 | IR |
| Flockbots[1] | $500 | wheels | N/A | ∅18 | 3 | Wi-Fi |
| E-puck[1,2] | $850 | wheels | 13 | ∅7 | 1-10 | Bluetooth, ZigBee |
| Kobot[1,2] | $1174 | wheels | N/A | ∅12 | 10 | ZigBee |
| R-one[1,2] | $322 | wheels | 30 cm/s | ∅10 | 6 | RF, IR |
| ZeeRO[1] | N/A | wheels | N/A | ∅25 | N/A | Bluetooth |
| MinDART[1] | N/A | tracks | 17 | 29x24x37 | N/A | None |
| marXbot[1,2] | N/A | treels | 50 | ∅17 | 4-7 | Wi-Fi |
| JL-2[1] | N/A | tracks | 20 | 35x25x15 | 2 | Wi-Fi |
| Khepera IV[1, 2] | $2700 | wheels | 100 | ∅14 | 4-7 | Wi-Fi, Bluetooth |
One of the first questions, coming up in the beginning of a project concerning swarm robotics, is money. Conducting real experiments with dozens of robots require high investments to buy all needed hardware and software. The price of the robots varies greatly that depends on a number of sensors and actuators, quality of a robot's frame, and demand for a particular robot. Unfortunately, the price of some robots is not specified. The most of these robots have an open-source design and code so they can be built from scratch in case there is such a need.
The table shows that the majority of the robots are differential drive robots whose movement is based on two wheels placed on either side of the robot. It allows achieving higher speed and precisely controlling the direction of movement. However, there are also some disadvantages, such as a need for a prepared surface without pits and bumps. Robots using the tracks usually can move on the unprepared surface, which facilitates a preparation for experiments and demonstrations. A unique type of locomotion is vibration (check the Kilobot robot) which tends to be slow and imprecise way of movement. It also requires a special smooth surface.
Size matters in the field of swarm robotics. A swarm consists of a large number of individuals which are quite effective in performing of important tasks for surviving of the colony. A significant number of robots is important in order to reproduce the swarm behavior or to check a hypothesis.
Unlike individuals in the real swarms, robots cannot generate energy from organics and still require a source of energy such as a battery. The working time of a robot depends on the capacity of battery and the amount of sensor and actuators using by a robot in experiments. Some of the robots have charging stations (e.g. Elisa, Jasmine) which may be useful in conducting of long-term experiments. The marXbot goes even further supporting exchange of the battery in less than 10 seconds without shutting down the power.
Communication in swarm robotics can play a crucial role while the effectiveness of the whole colony depends on how well the robots can send and receive messages. There are two types of communication: abstract (e.g. Wi-Fi, Bluetooth, ZigBee, RF) and situated (e.g. IR). Both of them work well for messages exchange, but situated communication provides additional information about the sender and the message, e.g. strength of the signal and its direction.
The choice of the robot depends on experiments where it will be used. Select the most important features of the future research and pick a robot that fits best.
No comments:
Post a Comment