+1-817-968-5551
+61-488-839-671
+44-7480-542904About Expert
Key Topics
Requirement
Write 7 pages essay on Robotics: past, present, and future
Robotics: the past, present, and future
Solution
Introduction
This research paper deals with the study of past, present and future of robots. The paper starts by discussing the origins of the robots and where the study is undergoing along with providing insights regarding the future of robotics. Thus, the paper covers the fundamental conceptual structures as well as discusses the potential fertile application domains.
Past Robotics
Practical robotics had its origins inside factories manufacturing products on assembly lines, where speed, precision, and reliability were paramount. Thus, both human tedium and imprecision were done away with while improving the uniformity of a quality product which could come off the assembly line relentlessly around the clock. Precision machine tools are considered the inspiration for such a development. However, before robots could be programmed using a computer, it was not possible to easily change any detail of their repetitive function. With the flexibility of programming and the training supplied by a skilled operator, robots could carry out very complex, humanlike, repetitive tasks; however the task could be changed at short notice by using a different program. Coping with environmental variability could only be accomplished with sensor feedback, this leading to what was known as the third robot generation. It soon became apparent that higher degrees of intelligence were needed to accommodate variability and uncertainty. As robots migrated out of the fixed automation, fully structured factory assembly lines into the unstructured and unpredictable worlds of space, underwater, in the air and on the ground, where many of the future applications could be imagined, it became clear that a complementary range of sensors and considerable artificial intelligence would be needed to achieve autonomy. This striving for autonomy in complex, unstructured and unpredictable environments, sometimes cohabited by humans became and still is the holly grail of robotics and has given rise to the field of 'intelligent robotics' where perception, reasoning, and actuation are highly coupled to achieve useful tasks with little human guidance [See Figure 1]. The pursuit of this goal, while promising practical outcomes, is also laced with no small measure of romance, in the idea of fabricating a human-like 'creature' of intelligence and grace which can be a human helper and companion.
Writing Assignment and completing it on time is not an easy job. If you forget to write your assignment you can take urgent essay help from Allassignmenthelp.com. We are a team of professional assignment writers who are talented enough to deliver a plagiarism-free assignment within the given frame. On our website, the students can get information technology assignment help at a very affordable price.
Present Robotics
Now we have access to a very rich set of sophisticated sensors, powerful computing platforms and all variety of agile mechatronic devices, it would seem that nothing should hold us back in achieving the dream of a fully autonomous robot agent which could carry out a variety of complex tasks in unstructured environments and at the same time be able to interact cooperatively with humans. However, this is still far from being achieved. Considerable advances have nevertheless been made. It is useful to consider the basic ingredients of a fully fledged robotic system at a rudimentary (functionality) level. First, there is a need to know how a robotic manipulator's links are configured in space or where a mobile robot is [Jarvis,1993] and how it is posed. Since manipulator links are interconnected and the geometry is known, a combination of shaft encoder readings (or equivalent) and the derivation of the inverse kinematics can determine the position and pose of the end effector (hand) fairly accurately. Flexure, backlash, and pay-load variations do, however, add the error, but this is usually manageable, perhaps with the addition of special devices such as lasers and cameras. Attempts of carrying out similar evaluations on a free-roaming mobile robot using the only odometry have not been satisfactory since wheel slippage, load distribution, support surface undulations, and wheel shape imperfections introduce errors which are accumulative and unrecoverable from. Following fixed lines on or signals under the floor, while workable, restricts the mobility of the robot. The use of beacons at known locations and scanners [See Figure 2] can be very effective, but site preparation is still required. There has been a definite trend in recent times to use only natural landmarks to achieve location/pose (localisation) without site preparation, sometimes without the even knowledge of the site beforehand. If the site is known, sensors on-board the robot still have to perform the 'data association' task to relate current sensor measurements to previously stored map information. Usually, continuity and speed constraints, together with odometric, visual, radar, range (ultrasound or laser based) or inertial information, can simplify this task. However, if the robot is picked up and deposited in a new place without the system's knowledge, determining the robots localisation (the so-called 'kidnapped robot' problem [Spero, 2006] is more difficult and global localisation strategies are required. There are essentially two modalities of landmark natural exploitation for robot localisation, one depending on specific, selected landmarks, and the other on the entire 'shape' of the sensed environment.
The precise landmarks must be unique, temporally stable, detectable, and visually stable, distributed and vertically compact. While comparing the live sensor data that was acquired on-board along with a known environment, the problem related to the correspondence matching problem should be resolved with the minimum amount of ambiguity and with accuracy and reliability. Critical localization can be caused due to mismatches but the constraints like continuity and velocity can also be utilized for detecting inconsistencies or for reducing ambiguity towards a satisfactory level. The second modality uses the whole set of live sensor data with high resolution and vision for estimating where these robots must be present for making these observations by corresponding to the environmental map database. Usually, a pure appearance based sensor data is employed. [Jarvis et al., 2007]
In the literature, a wide variety of matching techniques has been observed. The particle filters and their applications [Fox et al.,2001], is a type of genetic search algorithm in which the randomly chosen hypothesized places are analyzed and tested against the matching criterion and the resampling i.e. a probabilistic methodology further narrows down in the favour of the matches only until compact distribution of the feasible output are obtained. The odometric as well as inertial sensor data are used for predicting relative realization for allowing the drifting of these particles for better hypothesis positions and for reducing the computational search costs. A dilemma might be caused when the attempts for localizing without the formal environmental map would be made. When the position is known to the robot, it can easily collect the environmental sensor data and can also save a map on the position. But if the robot moves for collecting the new data, then the matching with already saved data, the robot can estimate a new position but can introduce some problems. The new data will be saved with some error. This is known as 'simultaneous mapping and localisation’ (SLAM) problem [Dissanayaka et al.,2001] and has generated a lot of publicity and is now considered as a centerpiece of effort amongst the robotics research community for over a decade years.
The laser range finders are also seen to suffer from specularity but can only range over thousands of meters. Most common laser range finders are seen in the robotics lab (figure 3), that includes Erwin Sick, Hokuyu, and Riegl. The riegl laser range scanners can also have built-in high-resolution digital cameras for producing color rendered 3D maps along with high ranges of 800m with accuracy. Only those surfaces can be scanned that are visible, and even multiple numbers of scans can be taken from various points for getting a full 3D map.
When the issues of localization and the mapping are clear there is a dual requirement towards the planning of path and the avoidance of obstacles forthe mobile robots for navigating from a point to a selected goal. Various path planners like Rapidly Exploring Random Trees (RRT) [LaValle and Kuffner,2001], a representable set is the A* methodology [Lozano-Perez and Wesley,1979], and Distance Transforms(DT) [Jarvis,1994]. All these planners have pros and cons. The DT has advantages that help it to accommodate initially not known environment, alterations in terrain navigability, explorations of unknown spaces, and covers all the available territory and planning of the overt paths that are hidden from some of the vantage points.
It has the capacity of extending various dimensions. It can also operate in tessellated spaces and can remain memory hungry when specifically high resolution has to be covered. The RRT and A* have some disadvantages but are useful in real Euclidean spaces and for better geometric than based on area.
Some examples are given in figure 4. Some of the presently studied robotics are Bipedal Gait Humanoid (Figure 5), Wakamaru Service Robots (Figure 6.) and Robot Swarm (Figure 7)
Future Robotics
Leaving Science Fiction aside, the expectations concerning intelligent robotic technology development over the next decade or so are quite modest. The practical application domains where robotic technology is most likely to be used are:
-
1. Transport (public and private)
-
2. Exploration (oceans, space, deserts, etc.)
-
3. Mining (dangerous environments)
-
4. Civil Defense (search and rescue, fire-fighting, etc.)
-
5. Security/Surveillance (patrol, observation, and intervention)
-
6. Domestic Services (cleaning etc.)
-
7. Entertainment (robotic toys etc.)
-
8. Assistive Technologies (support for the fragile)
-
9. War Machines
-
10. Scientific Instrumentation (e.g. synchrotron sample preparation, chemical screening, etc.)
Since the last of these is more like an assembly line process, it has been one of the first to have already been deployed commercially. Complex aspects, such as the capture and mounting crystals for x-ray crystallographic analysis, have not yet been automated. All the other domains involve high degrees of unstructured-ness, where commensurate levels of artificial intelligence based on sensor data fusion and understanding are required. An interesting way to approach autonomous operation whilst realizing useful applications along the way is to devise the means by which humans can interact and intervene [See Figure 8.] with robots which are richly sensor equipped, providing the missing capabilities such as subtle judgments, risk analysis, fine dexterity and reaction to unpredicted events but in such a way that a continuum between full autonomy and full teleoperation can be smoothly engaged. As the technology matures, the human support can be gracefully withdrawn with less and less intervention over more and more of the tasks. For example there are many situations in, say, firefighting where a robotic vehicle carrying water may move along a fire front spraying water at hot spots detected using a thermal camera fairly autonomously, but a human may need to direct the vehicle to move to another more critical, fire front or assist in a delicate rescue mission under direct the tele-operational control. As another example, a transport vehicle may safely negotiate a highway navigation task, changing lanes and speed as required, re-planning routes and so on, yet a human operator may need to take over at an unexpected construction site or scene of an accident.
Place Order For A Top Grade Assignment Now
We have some amazing discount offers running for the students
Place Your OrderSome new robotics research interests are developing in the bio-robotics and cognitive science fields, where the melding of biology with robotics has exposed new insights into the functions of mammalian brains, pathways to muscle functions and perception and has led the way towards evolving new prosthesis devices for both mechanical and perceptive needs. While biologists and cognitive scientists are using robotic notions to help them develop models of physiological function and test them in robotic experimental ways, so also have roboticists been inspired by deeper understandings of biological processes in their design of robots [Beer et al.,1997]. Thus the future for robots being used at various and variable levels of autonomy in many everyday activities is likely to be bright if one does not expect too much. The more exciting challenges will be where humans and robots work together cooperatively with a need for reliable communication and due regard for safety and reliability. Probably only mass production will permit the domestic robots to be affordable by most in much the same way as automobiles have proliferated. It is likely that affordable computational power will continue to grow, perhaps doubling in processing capability every two years at the same cost as it has for some time already. Sensor technology will not only continue to improve in capability, but its price will rapidly decline. Initial navigation systems which cost tended of thousands of dollars a decade ago are now available for a few thousand dollars or less. The development of robust methodologies for navigation, recognition and human interaction will continue, with modest improvements generated incrementally rather than in leaps and bounds. It is clearly here where the most research effort is needed. The melding of computational and robotic sciences will be essential for these improvements. One somewhat frightening possibility may impinge upon the development of personal robots to do chores for humans. The fear that genetically engineered sub-humans may take over such roles is certainly worth contemplating but preferably not embracing. Imagine a world in which being much like ourselves are happy to do our chores for very little reward or freedom. Where then would robotics still be relevant? That such developments are currently regarded as unethical and immoral may not be sufficient deterrent for such developments, perhaps initially by the unscrupulous, but may be later adopted more generally. Perhaps mere ignorance has raised this concern. Surely heavy duty operations in mines and industry, mineral exploration, hazardous environments and a host of other applications of this kind would still be robot related, but the need for sophisticated machine intelligence could be severely diminished. The main challenges for the future of intelligent robotics are:
-
1. Enhancements of the quality, smaller size, robustness and low price of the camera, ultrasonics, laser range, radar, and inertial sensors.
-
2. Enhancements in computational power at reduced cost. This feature will not want any singularconsiderationdue to existing market forces.
-
3. Enhancement in mechanisms for robot platforms is it capability, weight, strength, and the use of new materials, which includes ceramics, titanium, carbon fiber, etc.
-
4. Enhancements in navigation algorithms that includes natural landmark based approaches, accommodation of changing cost structures that are related to the navigability,recovery mechanisms collision risk, visibility, etc.
-
5. Enhancements of Human/Machine cooperation that includes task refinement, communication, intervention, etc.
-
6. Enhancement in risk assessment as well as endurance regarding graceful degradation and operational times.
-
7. Explaining of legal characteristics of humans and robots working together.
-
8. Improved understanding of emotional features of robots was working with humans.
-
9. Evolution of the robot/biology cross-inspirational trend.
-
10. Development of robotic ethics.
Conclusions
The vision for the development of the robotics seems positive, but the whole idea is evolutionary rather than revolutionary, along with a constant penetration towards the domestic and industrial world along with affordable prices. When the prices per unit decline via the mass production exactly like the case of personal computers and automobiles, then it would be observed that the simple robotic devices will replace the standard peripherals like printers. This would help in generation of intelligent robotics built in-washing machines, automobiles, lawn mowers and entertainment systems. The process is believed to start as a trickle, but soon after its success it is expected to deluge proportions, and it is expected that the process, just like any innovative technology being mind-blowing once becomes common later, would become affordable and compliant with the legal systems. Along with being a crucial paradigm shift, the main driving force of the process would be the substitution of sensor-based intelligence for high-speed and accuracy for accommodating unstructured-ness and uncertainty.
Reference
-
Jarvis, R. (n.d.). INTELLIGENT ROBOTICS: PAST, PRESENT AND FUTURE. International Journal of Computer Science and Applications, 5(3).
-
Jarvis, R.A.(1993) A Selective Survey of Localisation Methodology for Autonomous Mobile Robot Navigation, accepted for presentation at the Robots for Competitive Industry Conference, July 14-16, Brisbane Australia, pp. 310-317
-
Spero, D., (2006) Simultaneous Localisation and Map Building: The Kidnapped Way.
-
Jarvis,R.A., Ho, N. and Byrne, J.B,(2007) Autonomous Robot navigation in Cyber and Real Worlds, CyberWorlds 2007, Hanover, Germany, Oct. 24th to 27th, pp. 66-73.
-
Fox, D., Thrun, S., Burgard, W, and Dellaert, F.(2001) Particle filters for mobile robot localization. In Doucet, A., de Freitas, N. and Gordon, N. editors, Sequential Monte Carlo Methods in Practice, pages 499-516. Springer Verlag
-
Dissanayaka, M., Newman, P., Clark, S., Durrant-Whyte, H., and Csorba, M.(2001) A solution to the simultaneous localisation and map building problem. IEEE Trans. on Robotics and Automation 17, 3 June , 229-241.
-
LaValle, S.M. and Kuffner.(2001), J.J, Rapidly-exploring random trees: Progress and prospects. In B. R. Donald, K. M. Lynch, and D. Rus, editors, Algorithmic and Computational Robotics: New Directions, pages 293--308. A K Peters, Wellesley, MA,
-
Lozano-Pérez, T., and Wesley, M.A.,(1979) An Algorithm for Planning Collision-Free Paths Among Polyhedral Obstacles. Commun. ACM 22(10): 560-570
-
Beer, Randall D., Roger D. Quinn, Hillel J. Chiel, and Roy E. Ritzmann.(1997) Biologically Inspired Approaches to Robotics. Communications of the ACM, March : 30-38.
Give your grades a boost
Just share your requirements and get customized solutions on time.
-
1,168,768 Orders
-
4.9/5 Overall Rating
-
5,052 Experts
