Bowling Assistant Mechatronics Circus & Mechatronics - - PowerPoint PPT Presentation

bowling assistant
SMART_READER_LITE
LIVE PREVIEW

Bowling Assistant Mechatronics Circus & Mechatronics - - PowerPoint PPT Presentation

Bowling Assistant Mechatronics Circus & Mechatronics Symposium 2020 Sampo Laine Topias Turunen Martin Guggemos Contents How do we tell a good bowling performance? Key points How does the system work? IMU, Sensor fusion


slide-1
SLIDE 1

Bowling Assistant

Mechatronics Circus & Mechatronics Symposium 2020

Sampo Laine Topias Turunen Martin Guggemos

slide-2
SLIDE 2

Contents

  • How do we tell a good bowling performance? – Key points
  • How does the system work? – IMU, Sensor fusion
  • How do you receive feedback? – Web Interface
  • How precise is the system? – Reference device, Validation
  • What is the takeaway? - Discussion
slide-3
SLIDE 3

Some key aspects in bowling

  • Swing plane
  • Step timings
  • Ball-Ankle distance (5)
slide-4
SLIDE 4

Sensor unit ESP32 microcontroller Core 0 Web service hosting Core 1 I2C communication with IMU sensors Mobile device browser instance WIFI Data transmission Data processing and preformance calculation Web user interface SD-card module Temporary data storage IMU Digital motion processing

The application concept

slide-5
SLIDE 5

Sensor unit ESP32 microcontroller Core 0 Web service hosting Core 1 I2C communication with IMU sensors Mobile device browser instance WIFI Data transmission Data processing and preformance calculation Web user interface SD-card module Temporary data storage IMU Digital motion processing

slide-6
SLIDE 6

Wearable Sensor unit

Features:

  • Components hidden inside

small casing

  • Wireless measurement
  • Modular design allows for

multiple configurations

slide-7
SLIDE 7

Sensor unit ESP32 microcontroller Core 0 Web service hosting Core 1 I2C communication with IMU sensors Mobile device browser instance WIFI Data transmission Data processing and preformance calculation Web user interface SD-card module Temporary data storage IMU Digital motion processing

slide-8
SLIDE 8

Inertial Measurement Unit

  • Multiple sources of data
  • Signal drifting
  • Data accuracy

Things to worry about:

slide-9
SLIDE 9

Sensor fusion

  • The measurement data of

from one sensor is insufficient to measure the

  • rientation of the human

body

  • Combination of multiple

sources of data is required

slide-10
SLIDE 10

Sensor unit ESP32 microcontroller Core 0 Web service hosting Core 1 I2C communication with IMU sensors Mobile device browser instance WIFI Data transmission Data processing and preformance calculation Web user interface SD-card module Temporary data storage IMU Digital motion processing

The application concept

slide-11
SLIDE 11

User Interface

  • Web user interface
  • Performance consuming

calculations can be done on the mobile device

  • Data from bowling the

performance is presented

  • Multiple attempts can be stored

Instant feedback –

slide-12
SLIDE 12

Reference Device

α, 1. DOF IMU

  • Two degrees of freedom

pendulum resembling the elbow joint of a person

  • Two rotary encoders

capture the reference position

IMU Axis 1 Axis 2

slide-13
SLIDE 13

Validation measurements

  • Comparison between encoder and IMU rotation

data shows good correlation

Time (s) Angle (deg)

slide-14
SLIDE 14
  • The system is accurate enough to determine the

movement of a human body

  • Various performance characteristics can be extracted
  • Accurate feedback can be given to user
  • More motion tracking use-cases are conceivable

Discussion

slide-15
SLIDE 15

Any Questions?

slide-16
SLIDE 16

Thank you!