TY - GEN
T1 - Bringing IoT to sports analytics
AU - Gowda, Mahanth
AU - Dhekne, Ashutosh
AU - Shen, Sheng
AU - Choudhury, Romit Roy
AU - Yang, Xue
AU - Yang, Lei
AU - Golwalkar, Suresh
AU - Essanian, Alexander
N1 - Funding Information:
We sincerely thank our shepherd Dr. Anirudh Badam and the anonymous reviewers for their valuable feedback. We are also grateful to NSF (CNS - 1423455) for partially funding the research. We acknowledge the support of various teams in bringing the system together. 1) Hardware and Software engineers at Intel[11] for helping on various design aspects of the embedded sensor hardware and software. 2) D2M[3] for ball design and prototyping. 3) The International Cricket Council (ICC) [12] and Narayan Sundararajan for providing domain expertise on the game of cricket and the business of sports analytics.
PY - 2017
Y1 - 2017
N2 - This paper explores the possibility of bringing IoT to sports analytics, particularly to the game of Cricket. We develop solutions to track a ball’s 3D trajectory and spin with inexpensive sensors and radios embedded in the ball. Unique challenges arise rendering existing localization and motion tracking solutions inadequate. Our system, iBall, mitigates these problems by fusing disparate sources of partial information – wireless, inertial sensing, and motion models – into a non-linear error minimization framework. Measured against a mm-level ground truth, the median ball location error is at 8cm while rotational error remains below 12◦ even at the end of the flight. The results do not rely on any calibration or training, hence we expect the core techniques to extend to other sports like baseball, with some domain-specific modifications.
AB - This paper explores the possibility of bringing IoT to sports analytics, particularly to the game of Cricket. We develop solutions to track a ball’s 3D trajectory and spin with inexpensive sensors and radios embedded in the ball. Unique challenges arise rendering existing localization and motion tracking solutions inadequate. Our system, iBall, mitigates these problems by fusing disparate sources of partial information – wireless, inertial sensing, and motion models – into a non-linear error minimization framework. Measured against a mm-level ground truth, the median ball location error is at 8cm while rotational error remains below 12◦ even at the end of the flight. The results do not rely on any calibration or training, hence we expect the core techniques to extend to other sports like baseball, with some domain-specific modifications.
UR - http://www.scopus.com/inward/record.url?scp=85076910963&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85076910963&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:85076910963
T3 - Proceedings of the 14th USENIX Symposium on Networked Systems Design and Implementation, NSDI 2017
SP - 499
EP - 513
BT - Proceedings of the 14th USENIX Symposium on Networked Systems Design and Implementation, NSDI 2017
PB - USENIX Association
T2 - 14th USENIX Symposium on Networked Systems Design and Implementation, NSDI 2017
Y2 - 27 March 2017 through 29 March 2017
ER -