Power Consumption and Conservation of Industrial Robots Through Embedded System

INTRODUCTION: 
 1) Shutdown of Unused Components: Electric
Robot was coined by Czech playwright Karl Capek incomponents consume static power in idle states.
his play R.U.R (Rossum's Universal Robots), whichShutting down the power supply when a component is
opened in Prague in 1921. Robot is the Czech word foridle can save the static power. When the robot stops,
forced labor.the sensors may be turned off. If half of the time the
The term robotics was introduced by writer Isaacsensors can be shut down, the average sensing
Asimov. In his science fiction book I, Robot, published inpower can be reduced.
1950, he presented three laws of robotics: 
1. A robot may not injure a human being, or, through2) Sensing Frequency Scaling: It is intuitive that the
inaction, allow a human being to come to harmsensing frequency should be different when robots
2. A robot must obey the orders given it by humanmove at different speeds. The sensing frequency
beings except where such orders would conflict withneeds to be higher when the speed is higher. Instead
the First Law.of keeping the sensing frequency that satisfies the
3. A robot must protect its own existence as long ashighest speed’s need, we can reduce the sensing
such protection does not conflict with the First orfrequency when the robot moves slowly. If the robot
Second Lawmoves slowly and the sensing frequency can be
 reduced.
  
 3) Dynamic Voltage Scaling: DVS is very effective in
INDUSTRIAL ROBOTS:reducing processors’ power. The processor inside
 the Hitachi-8s microcontroller can work at two
Robots usually have multiple components, such asdifferent frequencies: 20MHz and 10MHz. The current
motors, sensors, microcontrollers and embeddedoperating system inside the microcontroller doesn’t
computers. DC motors transform direct current intosupport the frequency scaling. Therefore, we can not
mechanical energy and are often used to drive themeasure the power savings. However, if we can
robots. Sensors collect data from environment anddynamically change the working frequency according
provide information to robots. Most often used sensorsto the workload, we can reduce the control power.
are vision, infrared, sonar and laser rangers. ManyThis technique also applies to the embedded computer.
robots use embedded computers for high-level 
computation and microcontrollers for low-level controls.4) Trade-off between Motion and Communication:
 A Team of robots may move and cooperatively
 execute a task. Robots need to send sensing data
 through wireless communication. Consider one robot
 needs to transfer data to another robot, but the robot
 is far away. If the robots can move closer, the
 communication power can be saved. The cost here is
EPSON INDUSTRIAL ROBOTthe motion power for moving closer. If the volume of
 the data is large enough, more communication power
 can be saved than the motion power cost.
  
 5) Energy-Efficient Real-Time Scheduling for Robots:
 A mobile robot is a real-time system. The robot can
 have many periodic tasks, such as motor and sensor
 The microcontroller directly controls motors, sensors,control, sensing data reading, motion planning, and data
and polls the sensor readings. It hides the hardwareprocessing. The robot may also have some aperiodic
details from the embedded computer, and provides antasks, such as obstacle avoidance and communication.
application programming interface (API) for theRTS can work with DPM to more effectively reduce
embedded computer. The embedded computerthe power consumption. For example, if a scheduler
handles high-level computation, including motion planning,can cluster tasks closer in time and create longer idle
image processing, and scheduling. The separation ofperiods, shutdown techniques can be more effective.
the microcontroller and embedded computer makesRTS also can work with DVS to reduce processor
the designs more flexible. However, other componentsenergy consumption, as we discussed in the related
like sensing, control, communication and computationwork. For mobile robots, the tasks’ deadlines are
also consume significant amounts of power. It isdifferent at different traveling speeds. At a higher
important to consider all components to achieve betterspeed, the periodic tasks have shorter periods.
energy efficiency. This study has two majorTherefore, we should consider both motion planning
contributions. Firstly, we study power consumption of aand RTS together.
robot 
 FORTIFICATION:
USES OF INDUSTRIAL ROBOTS: 
            One of the most common usesMany fails to consider the third quadrant called
for industrial robots is welding. Robot welded carfortification. This paper put some idea about that
bodies for example enhances safety, a robot neverquadrant. Some shrinks the use of robots by Fleet
miss a welding spot and performs equally all throughSize Problem [3]: A fundamental question for
the day.multi-robot applications is to decide the number of
In assembling of parts many of these robots can berobots needed (i.e., the “fleet-size problem”) to
found in the automotive and electronics industriesaccomplish tasks. We provide a probabilistic method to
Packaging/palletizing, is still a minor application area fordecide the fleet size necessary to serve requests with
industrial robots, this application area is expected torandom arrival times and locations. We consider five
grow as robots become easier to handle.factors on which the fleet size depends: available
The food industry is an area where robots areenergy, power consumption, service field, request rate,
expected to play a major role in the future. Theand timing constraints.
process involves harvesting each arriving plant, cutting 
its steam into segments near each node, and thenThough we know that many of the industrial robots
replanting the segments so that they can grow intouses stepper motors, servo motors, relays etc., on AC
new plants etc.or DC. In AC supply it is must to keep the power
 factor under control it should not be low. If it so, heavy
 MICROCONTROLLER AND EMBEDDEDpower loss will be occurred and the company will be
   COMPUTERliable to meet the surcharge of their electricity
 department.. The following are the some of the
The microcontroller periodically sends commands tofortification technique
motors and sensors, polls sensors’ readings, and 
communicates with the embedded computer. The1. To improve the power factor capacitor has to be
microcontroller’s tasks are usually fixed so theused. Though robot uses electrical motors,
power consumption of the microcontroller can betransformers etc., under starting state it needs high
modeled by a constant. The embedded computer iscapacitance to maintain pf; on the other hand under
more complex than the microcontroller. Many studiesrunning condition it needs minimum capacitance value. In
have been devoted into simulation-based methods toother words it can be explained as the capacitance
estimate its power consumption [6] [5] [8]. The powervalue changes according to the load of the robots.
consumption of the embedded computer may varyThese will be an extra burden to cutoff the capacitor
significantly across different programs.under load. To overcome this dynamic control
capacitor may be used.
2. The one of the factor that decides the life of the
 robot is wear and tear of the electrical equipments.
 This falls in sparking of the relays contacts and motor
 brushes due to the in rush current. These can be
Micro controllereliminated by close circuit transient.
 
 3. Due to energy conservation law, energy will be
dissipated through heat. This can be reduced by silver
 windings. Silver has the lower resistance of current
 then copper and aluminum.  Due to lower resistance
 of electric current it reduces major electrical losses
Motor 
4. Over load detector for example, robot with
Sensorhardware platform of a Pioneer 2-DX robot [2]
Embedded computeraugmented with custom hardware for watering. To
deliver water to the plant, the robot has been fitted
 with a water line, dispensing spout, and pump. To
 deliver power to wireless sensors an inductive
 charging coil has been positioned near the watering
PREVIOUS WORK:spout. Similarly, another paddle shaped inductive charge
 coil has been added to the robot to allow it to
[1] Both timing and energy constraints are considered;recharge itself at its “maintenance bay”. In order
the robots carry limited energy and need to finish theto support calibration, the robot includes a sensor node
tasks before deadlinesthat was human-calibrated lastly, the robot has a
 maintenance bay it uses to automatically charge its
 own batteries and refill its water reservoir. The reliability
ENERGY-CONSERVATION TECHNIQUES:of this approach has been demonstrated during the
 deployment of the robots Rhino and Minerva as
This section explains three promising techniques forautonomous museum tour guide robots [4, 7]. The
power reduction of mobile robots.high-level task ordering and dispatching software was
 custom-built for the Plant Care project.
A. Dynamic Power Management 
Dynamic power management (DPM) dynamicallyThere may be the chance to water to direct contact
adjusts power states of components adaptive to thewith the power pack. This leads to short circuit, and
task’s need. The purpose is to reduce the powerdraws more current that the rated (PU) per unit.
consumption without compromising system 
performance. Many electronic components haveFUTURE ENHANCEMENTS:
multiple power states; their power consumption is 
different at different power states. For example,For future work, I plan to extend the current study in
processors can run on different frequencies. To savetwo directions. First, we will measure power
power, the processors can enter lower frequenciesconsumption of more components, such as laser
when the workloads are light. Another example is torangers, cameras, servo motors, stepper motors, and
shut off the power supply to the disk in an embeddedrelays. Second, I plan to implement the proposed
computer to save the static power when there is noenergy conservation techniques into the Pioneer
disk access.robots, and conduct experiments in real applications
A simple DPM method shuts down a component 
when it is idle. It is essentially a prediction problem. If weCONCLUSION:
predict there is no access on this component for a 
reasonably long period of time, the component can beIn this study, I presented some of the power
shut down to save static power. Turning on and offconsumption technique of different components of an
the component takes time and energy. If the idle periodindustrial robot. In this paper, I introduce one technique
is too short, the components may actually consumecalled fortification technique than two exiting techniques
more energy for turning on and off. One of the widelyDPM and RTS for energy-efficient designs of robots.
used prediction methods is timeout: if the componentThese techniques together with motion planning
has been idle for a time period longer than the timeout,provide greater opportunities for reducing the power
the component will be shut down. The rationale behindconsumption and prolonging the operation time of
timeout is that the component is likely to keep idle inmobile robots.
the near future since it has been idle for a while. 
Another widely used DPM technique is dynamicREFERENCE:
voltage scaling (DVS) by reducing both supply voltage [1] Yongguo Mei, Student Member, IEEE, Yung-Hsiang
and clock frequency to reduce the powerLu, Member, IEEE, Y. Charlie Hu, Member, IEEE, and C.
consumption of processors. CMOS circuit is itsS. George Lee, Member, IEEE
dynamic power, which can be expressed by c Vdd, f, 
where c is the effective switched capacitance, vdd is[2].ActivMedia Robotics, visited Feb. 2002.
the supply voltage and f is the clock frequency. 
B. Real-Time Scheduling[3] Y. Mei, Y.-H. Lu, C. S. G. Lee, and Y. C. Hu.
Real-time systems handle tasks with deadlines.Determining the Fleet Size of Mobile Robots with
Real-time scheduling (RTS) schedules multiple tasksEnergy Cons traints. In IEEE /RSJ International
and meet the deadlines. If the tasks can be scheduledConference on Intelligent Robots and Systems, pages
without missing the deadlines, we say they are1420–1425, 2004.
schedulable. Mobile robots are real-time systems. 
When a robot detects an obstacle, it has to timely[4]. Burgard, W., A. Cremers, D. Fox, D. Haehnel, G.
slow down and decides the next motion. For multipleLakemeyer, D. Schulz, W. Steiner and S. Thrun. 1999.
robots coordinating to accomplish a task, timelyExperiences with an interactive museum tour-guide
information communicating is critical. Two often usedrobot. Artificial Intelligence.
scheduling algorithms are rate monotonic (RM) and 
earliest deadline first (EDF). Many other algorithms are[5] J. R. Lorch and A. J. Smith. Apple Macintosh’s
based on these two. RM is a fixed-priority algorithm,Energy Consumption. IEEE Micro, 18(6):54–63,
assigning a higher priority to a task with a shorterNovember 1998
period. EDF executes the task with the earliest 
deadline among all ready tasks. It has been proved[6] D. Brooks, V. Tiwari, and M. Martonosi. Wattch: A
that EDF is optimal with respect to minimizing theFramework for Architectural-level Power Analysis and
maximum lateness. Besides scheduling tasks to meetOptimizations. In International Symposium on Computer
their deadlines, RTS can also schedule the tasks suchArchitecture, pages 83– 94, 2000.
that DPM can save more energy. For example, when 
the idle periods of a component are too short due to[7]. S. Thrun, M. Bennewitz, W. Burgard, A. Cremers, F.
frequent accesses, power cannot be saved byDellaert, D. Fox, D. Haehnel, C. Rosenberg, N. Roy, J.
shutting down the component. However, if we canSchulte and D. Schulz. 1999. MINERVA: A second
reschedule the tasks and make the component havegeneration mobile tour-guide robot. In Proceedings of
more long idle periods, the component may be shutthe IEEE International Conference on Robotics and
down to save power.Automation (ICRA).
C. Examples 
In this section, we show some potential applications of[8] T. Simunic, L. Benini, and G. D. Micheli.
DPM and RTS into energy-efficient robot designs usingCycle-accurate Simulation of Energy Consumption in
several examples.Embedded Systems.