June
05
0 likes
2 comments
0 shares

The Slow Trudging Advance of Robotics

As time marches on, so does the presence and ability of machines and computers marching with us. Many people fear the day when computers and machines begin to walk on their own, but whether it is due to science fiction or actual issues present in the progressing system, they still have a far way to go before then.

Despite the distance to these possibilities, many people wish to push the envelope further and bring self sufficient robotics around even sooner. Among the dozens of robotics competitions held annually all over the world there are many that stand out. One that is particularly eye catching is that of the DARPA Robotics Challenge.

This challenge is specifically meant to usher developers towards a point where Robots can act and work without the need for human instruction, at least in the perspective of a natural disaster and to help in relief efforts. The challenge takes robots and presents them a series of tasks that are complex, require a certain level of strength, demonstrate the dexterity of the robot, all while losing and regaining connection with the controllers.

Such tasks include driving a vehicle, maneuvering through a loose and uncertain landscape of rubble, opening doors, and even cutting through a barrier made of drywall. Such tasks require a very diversely equipped robot that can go beyond much of the current level of analytics.

With many contenders it is a jam packed event as many try their hands at the challenge. So long as the robot can complete the challenges then it can be considered a success. There is no cheating in the contest as that anything that can give the robot a boost over the competition is the result of a lot of work by the engineers and designers to give it said advantage. Some robots are able to switch between legs and four legged tread like transportation in order to cross over areas that are likely to cause a fall. Others are solely bipedal and require extensive calibration in order to handle uneven terrain. The many designs and approaches to making it through the competition are all vying for a $2 Million prize in funding to continue with their project.

The DARPA Robotics challenge is a great initiative to press the field for constant advancement. The DARPA event was actually spawned as a result of the Fukushima Disaster in 2011, When trying to survey the results of the disaster from a remote location, time after time robots were being found to be ineffective or, at the worst, completely unequipped to deal with the challenges that were presented to them.

From simple tasks like turning door knobs to moving through bad terrain, the robots used were failing to meet the challenge left and right. In order to push the frontier and presence of robotics in addition to their ability to be used in disaster relief, the DARPA challenge was created. Some engineers compared the level of intelligence now displayed by some of the top robots being tested lies between that of an insect to that of a small mouse. While this may not be very much, it shows that the field of robotics has made an immense movement forward in the time that it has been developing.

To prepare for the finals, running today and tomorrow, each team was allotted $1.5 Million in funding as well. Getting an hour to run through the entire course and score as many points as possible, all the teams will be competing simultaneously. However slow of an event that it might end up being (Compared by some to paint drying) it is an important step in the future of robotics.

Share:

2 Comments

  1. Zhen Li on June 8, 2015 at 12:19 pm

    It’s interesting you mentioned “Some engineers compared the level of intelligence now displayed by some of the top robots…” My question is, “how are they actually measuring the intelligence,” and how does that compare to the recently launched Japanese robot that they set up in a grocery store Aico Chihira (They gave it a real name…):

    http://www.roboticstrends.com/article/meet_chihiraaico_the_creepiest_robot_of_ces_2015

    • Gavin on June 8, 2015 at 12:45 pm

      In the same fashion that intelligence has been hard to concretely label in humans, it is hard to label in robotics as well. Some things that they become measured on include adaptability, autonomy, and understanding. While certain topics such as mathematics or the location of different objects are topics with more definitive answers, topics like philosophy or psychology are harder to work with. This understood, different robots are created with different intentions and this can mean they are intelligent in ways that do not correlate. Aico Chihira, from what I have seen, seems to display a fair understanding of processing information that she is given so that when she is asked for directions, prices, or other information, she can easily give it. As I found also, the engineers working on her are still working on how she reacts to different stimuli. In this factor she is definitely more emotionally intelligent then other robots. If placed against some of the robots in the DARPA challenge I mentioned, those robots would be able to surpass Aico Chihira on how to intelligently move through a field of debris or provide support in dangerous emergency situations. You do indeed have a valid point, though, on the ambiguity displayed in terming robots with ‘Intelligence’ so far as it can be measured at the current point in time. It seems that they should be measured more in terms of capability then that of intelligence.

LEAVE A COMMENT