You are here
An artificial intelligence based third Umpire
The “gentleman’s game,” cricket, has adopted technology to assist on-field umpires. But it is used by the third umpire who is off-field and whose decision is overriding.
At his disposal are sensors, instruments and software that can give an indication as to whether and/or when the ball hits the bat, pads, stumps or part of the batsman’s body and its trajectory. There are also cameras in line with the popping creases at both ends, for run-out decisions. He is called upon to make a decision upon request for a review of the on-field umpire’s decision.
Considering the case of a batsmen who is given out for being caught by the wicket keeper. He is of the view that the ball did not hit his bat but rather his front pad as he drove at the ball with his bat away from his body.
In conducting the review, the third umpire starts to looks at a slow motion visual record/replay of the stroke. He tries to see if the ball did hit the bat. Many a times, because of the angle of the bat and the positions of the cameras, occlusion occurs and it is thus difficult to reach a conclusive decision. He then looks at “Hot Spot,” an infrared-based instrument.
Wherever the ball strikes, some heat energy is transferred, to that spot. This spot heats up and it can be detected. This will usually confirm if the ball did strike the bat, or the pad or both. To further confirm this, “Snicko,” which records the sound that occurs when the ball strikes a surface, is utilised.
When “Snicko” is used in conjunction with a slow motion replay, one can determine if the sound occurred when the ball was passing the bat or in fact when it was passing by the pad thus allowing for a determination of exactly where the ball struck.
Having viewed the outputs of the various instruments, the third umpire makes a decision which is then conveyed to the on-field umpire. The entire appeal and review process can take a minute or two and sometimes more. During this time the game is held up.
Additionally, some feel that the authority of the umpire may be undermined by this process as his original decision may be reversed upon receipt of the decision of the third umpire. In other words, he cedes his authority to the third umpire upon a review.
The use of an artificial-based decision system to assist the on-field to make his decision may be worth considering as it places the technology in his hands. The system could work like this. When an appeal is made, the presiding umpire who is outfitted with a transmitting and receiving system, sends a signal to the off-field based AI system. This system then automatically analyses (using image/pattern recognition techniques) the very same evidence that a third umpire would have access to and then determines if indeed the ball did strike the bat. This is then conveyed to the on-field presiding umpire.
Upon receiving the output (decision) from the computer-based “third umpire” he would be better equipped to make the best decision possible as he would have the benefit of a high tech decision support system in addition to his own expertise. This process would be faster and can certainly consolidate the authority of the on-field umpire as such a system would make the review procedure unnecessary. The need for a third umpire in this present set–up would also disappear.
Of course, post mortem reviews can provide useful information for training and deepening the expertise of umpires. The technology for developing a system like that described here exists and therefore the R&D associated with it is doable. Further, AI-based decision support systems are being applied with great success in a fields ranging from finance to security, warfare to medicine. Why not cricket?
User comments posted on this website are the sole views and opinions of the comment writer and are not representative of Guardian Media Limited or its staff.
Guardian Media Limited accepts no liability and will not be held accountable for user comments.
Guardian Media Limited reserves the right to remove, to edit or to censor any comments.
Any content which is considered unsuitable, unlawful or offensive, includes personal details, advertises or promotes products, services or websites or repeats previous comments will be removed.
User profiles registered through fake social media accounts may be deleted without notice.