technology-in-sport

Page 1

12

Scan

‘ That cobot was on the sensor!’ N

o one would want to argue with John McEnroe when he’s in full rant mode. The legendary tennis player was famous for winning many Grand Slam tournaments and for arguing with umpires and line judges. At one point during a match, he shouted at the umpire: “You cannot be serious! That ball was on the line!” The umpire and line judges said the ball landed outside the tennis court, meaning McEnroe would lose the point. McEnroe was livid. He walked up to the umpire and yelled at him. “You can’t be serious, man. You cannot be serious! That ball was on the line! Chalk flew up. It was clearly in. How can you possibly call that out?” The umpire stuck to his decision and McEnroe was fined for his outburst. To be fair, it wasn’t the first time McEnroe had had an argument with the court authorities. That was in the early 1980s, when even video replays were not much help in definitively deciding such matters. But nowadays, of course, most Grand Slam tennis tournaments, including Wimbledon, use artificial eyes to decide whether a ball was in or out. Players are given a number of opportunities to challenge a decision. The umpire will then replay the computer-generated image of the trajectory of the ball and leave it to the computer to show whether it really did land inside or outside the court. In the case of McEnroe’s infamous rant, the tennis player was more likely to have been correct than incorrect, according to new research. Having been criticised over it for so many years, it might mean something to him to be vindicated. But he has had a stellar career regardless, and it probably means more to him that he is still healthy enough to have fun playing tennis at masters tournaments around the world, even recently challenging Serena Williams to a match. Should the match ever take place, computer-generated replays would ensure that neither of the fiery-tempered players will have an excuse to harangue the officials. It’s an example of how people are willing to trust

editorial@roboticsandautomationnews.com

advertisement

Sensor Readings

Tech in sport Tennis was one of the first to adopt the Hawk-Eye system, but is the algorith behind it fair? And can the sport go even further to ensure no tennis player is reduced to an epic tirade of frustration over a bad line call

The Hawk-Eye system uses big data and mathematics to calculate the trajectory of the ball

computers and artificial vision more than human eyes for making decisions where there are fast-moving objects involved. It has always been the case that video replays can show things umpires and referees miss in sport. But now the technology is fully established. The vision system used in tennis is provided by a company called Hawk-Eye, which is owned by electronics giant Sony. It uses data captured from a range of cameras located around the tennis court, and calculates the trajectory of the ball, rather than show actual movements of the ball. It’s a mathematical model that is logical and sensible, but may not take account of tiny, sudden gusts of wind from nowhere that force the ball to move in mathematically unpredictable, or random, ways. Gusts of wind of that type – highly localised to a tennis ball and imperceptible to a camera – don’t exist in nature, but it’s one way to illustrate the point that no system is flawless. Moreover, the algorithm that Hawk-Eye is based on is assumed to be mathematically impartial – meaning, it does not favour any particular players. We assume this, as spectators, partly because we would not know what to check for even if the company was to show us all the code that goes into making Hawk-Eye. But there are some who would be able to, and at least one of them claims computer algorithms can be as biased as humans. Suresh Venkatasubramanian, an associate professor in the School of Computing at the University of Utah, led a team which devised a test to determine whether an algorithm was discriminatory and concluded that it would indeed be possible to detect bias in an algorithm. Venkatasubramanian and his team used a legal concept called “disparate impact” and demonstrated how it worked, as reported on TechRadar.com. Job applications were an area they examined. “The irony is that the more we design artificial intelligence technology that successfully mimics humans, the more that AI is learning in a way that we do, with all of our biases and limitations,” said Venkatasubramanian. “It would be ambitious and wonderful if what we did directly fed into better ways of doing hiring practices. But right now it’s a proof of concept.” Al Jazeera, the news channel, looked into Venkatasubramanian’s theories and made its own, far simpler contribution to the debate. A contributor to its tech show The Stream said: “Google ‘Beauty’ and see white women. Google ‘Asian’ and get porn. I think it’s as simple as that.” Venkatasubramanian said the problem is easy enough to fix by using his test to isolate the data that is creating the discrimination and then redistributing it so the bias disappears. He adds: “It’s more easy for these things [mistakes] to happen than one might imagine.” In tennis, it might be easy enough to resolve the problem by doing away with the chalk altogether and draw the court using sensors. Instead of lines made of chalk there would be lines made of thin sensor-laden material. Such material already exists and would, one imagines, provide definitive evidence on whether the ball was on the line or not, without needing to add code to account for random micro-storms. Taking tech in tennis one step further might involve coating the tennis ball with touch sensors which can interact with the sensors on the lines of the court. And then why not place some sensors on the racquet? And have the umpire wear Google Glass to augment his or her vision? Or even contact lenses with sensors in those? There are sports fans who are reluctant to allow too much technology in sport, and one can see their point. Screaming ‘ That cobot was on the sensor!” doesn’t quite have the same ring to it, and certainly would not echo in sports history like McEnroe’s tirade has done. l

www.roboticsandautomationnews.com

Offline programming Offline programming has never been easier thanks to RoboDK. You don’t need to learn brand-specific languages anymore. RoboDK handles the robot controller syntax and outputs the right program for your robot. Try a basic Pick and Place example.

Multiplatform RoboDK is the first multiplatform robot offline programming software. It works on Windows, Mac, Linux and Android devices. It even works on your phone or tablet! Check the download section.

Python powered RoboDK is a robot development kit that allows you to program any robot from any brand through Python. Python is easy to learn yet powerful and flexible. Robot offline programming has no limits with RoboDK’s Python API.

CNC friendly Use your robot like a CNC. Convert CAM files into robot programs, your robot can be used like a 5-axis CNC. Easily simulate the result with RoboDK and avoid collisions, robot singularities and joint limits. Download and try our robot milling example.

Robot accuracy Certificate robots. Check the accuracy of your robots with a ballbar test. Obtain a PDF report describing the accuracy and repeatability of your robots. RoboDK allows you to calibrate your robots and improve production results. Contact us for more information.

Extended library The RoboDK Library has many robots, external axes and tools from different brands. We are constantly adding new robots to RoboDK. The library can be directly accessed from our desktop app.

http://www.robodk.com


Turn static files into dynamic content formats.

Create a flipbook
Issuu converts static files into: digital portfolios, online yearbooks, online catalogs, digital photo albums and more. Sign up and create your flipbook.