1.
Architects Using Robots to Build Beautiful
Structures [http://spectrum.ieee.org/automaton/robotics/industrial-robots/robots-in-architecture]
2.
Teaching Robots to Think like People
http://www.gizmag.com/robots-machine-learning-cornell/19832/
The article on robotics from access science was a general
overview of different types of robots and how they work. Before reading the
article I had a general understanding of the types of robots and their thinking
process but after reading the access science article I learned more in depth about robotics. This was a good article to read before looking for additional articles on robotics.
The first additional article I looked at was very similar to
the video we saw in class. The robots in the article construct different shaped
structures based off of computer input. One issue with these robots is that the
size of the structure built is limited. The architects that designed these
robots are starting research on a robot that can construct a high-rise
building. This would mean mobilizing the robots. Some ways to mobilize them would be to put them on tracks or some sort of elevation lift. Some of the robots seen in the article cut the material being used precisely and then can construct
temporary spaces using the cut material. The
accuracy in which the machines cut the material is far better than that of human labor. The use of robots in construct is becoming a reality and this research is very import to develop the best robot for construction.
There are always issues that arise when a building is under
construction. A human knows how to problem solve to fix the issues that come up
but the robots in the previous article and the ones we saw in class are
programmed to complete a task and not solve a problem if something were to go
wrong. Researchers at Cornell are developing a robot that can think on its own in case there are any issues in construction.
For example, stacking plates in a dishwasher is an easy task if the plates are
the same size but a little more difficult when different. Their robot managed
to get 95% of these sorts of tasks correct when dealing with new objects in new
environments.This is a step forward to robots thinking somewhat on their own. There would still need to be people on the construction site to oversee the robots.
Both these articles discuss developments in intelligent building construction by using robots. The first article shows that it is possible for a robot to construct a building. Not only can the robot place the bricks or wood planks but it can also cut the material to the correct size. It also addresses some issues that arise when a robot constructs a building. One of these issues is that a robot is computer programmed to complete specific tasks. It cannot problem solve if something were to go wrong. The second article discusses this issue and what researchers are doing to start to solve this problem.
I think that in the very near future, machines will be able to perform basic problem solving, enabling it to maneuver a construction site. Through the use of sensors and communication between equipment, an accurate, realtime model of a construction site can be obtained. I feel this is the biggest obstacle facing automation of construction sites. I also believe that once automation of construction sites begin, sites will become more ideal for automated equipment. Currently, construction sites are organized so that average people can problem solve through obstacles. However, an automated construction site would become optimized for robotic equipment, allowing for less accidents and more productivity. I think the hardest hurdle to overcome would be a construction site consisting of half people and half automated equipment.
ReplyDeleteI think that robots will take an integral part in the development of buildings both during construction and operation. Like you mentioned robots are not able to problem solve and are very limited in the amount of tasks that they can perform before be reprogrammed or modified physically. I feel that the first issue of not being able to problem solve is not to big as long as the robots are able to determine when there is a problem. If they cannot and keep trying to work serious issues can occur. The second problem about being limited creates two main issues. If you choose not to have a person reprogram or retrofit the robots you need a lot of robots on site, taking up valuable space, especially in a city. Over all I think that robots will make a small appearance in construction in within the decade, but buildings will not be built without much human interaction for a long time.
ReplyDelete