Obviousness

An obviousness analysis under 35 U.S.C. § 103 requires identifying prior art that teaches or suggests the claimed invention, and providing a motivation for a person having ordinary skill in the art (PHOSITA) to combine those references to arrive at the invention.

Level of Ordinary Skill in the Art

A PHOSITA in the field of the invention (motion detection for computers, navigation, or portable electronic devices) at the priority date of January 6, 2010, would possess knowledge of various motion sensor technologies (accelerometers, gyroscopes, magnetometers), fundamental principles of rigid body dynamics, coordinate transformations (including quaternions), and established techniques for sensor fusion and signal processing used in inertial navigation systems (INS) or attitude and heading reference systems (AHRS). They would also be aware of the inherent limitations of individual sensors and simpler compensation methods.

Primary Prior Art Reference

The patent itself references and discusses several prior art patents by Liberty (U.S. Pat. Nos. 7,158,118, 7,262,760, and 7,414,611). Any of these can serve as a primary reference. For this analysis, we will refer to them collectively as "Liberty's prior art."

Liberty's prior art describes a pointing device that uses a "5-axis motion sensor" comprising three accelerometers (Ax, Ay, Az) and two gyro-sensors (ωY, ωZ) to detect rotation about the Yp and Zp axes. This device aims to detect motions and translate them to a cursor on a 2D display, including a compensation mechanism for signals affected by gravity or "roll" related rotations.

Differences Between Liberty's Prior Art and US11698687

US patent 11698687 distinguishes itself from Liberty's prior art in several key aspects, as explicitly stated in the patent text:

1. Sensor Modality (Nine-Axis vs. Five-Axis): US11698687 utilizes a "nine-axis motion sensor module" which includes a full three-axis gyroscope (detecting angular velocities ωx, ωy, ωz), a three-axis accelerometer (Ax, Ay, Az), and a three-axis magnetometer (Mx, My, Mz). In contrast, Liberty's prior art is limited to a "5-axis motion sensor" (three accelerometers and two gyroscopes detecting rotations about Yp and Zp axes).
2. 3D Absolute Orientation Tracking: US11698687 emphasizes its capability to "accurately obtaining and calculating actual deviation angles in the spatial pointer frame" (3D) and to output these in an "absolute manner" reflecting actual movements, excluding undesirable interferences. Liberty's prior art, according to US11698687, "may not output deviation angles... in a 3D reference frame but rather a 2D reference frame only" and yields a "planar pattern in 2D reference frame only." It also generates "relative" movement patterns, which can lead to errors at display boundaries.
3. Enhanced Sensor Fusion Algorithm: US11698687 describes an "enhanced comparison method and/or model" that involves representing device orientation using quaternions (e.g., previous, current, updated states), calculating "predicted axial accelerations" and "predicted magnetism" based on measured angular velocities, and then comparing these predictions with actual "measured axial accelerations" and "measured magnetism" to obtain an "updated state" (quaternion). This sophisticated process aims to "eliminate the accumulated errors as well as noises over time" and to specifically exclude "undesirable external interferences" such as non-gravitational accelerations and magnetic field disturbances. Liberty's compensation is described as simpler, mainly addressing gravity effects and lacking the robustness against dynamic interferences.
4. Data Association for Interference Rejection: US11698687 further incorporates a "data association model" to intelligently process sensor signals, such as by determining if measured states (e.g., accelerations or magnetism) are "good enough" to be used for compensation based on a "predetermined value or range," thereby allowing for selective compensation and rejection of corrupted data, for instance, due to electromagnetic fields.

Obviousness Combination and Motivation

A PHOSITA would have been motivated to combine the teachings of Liberty's prior art with other known technologies and techniques available by the priority date (January 6, 2010) to arrive at the claimed invention.

1. Motivation to include a 9-axis sensor module:

   • Liberty's prior art explicitly failed to output 3D deviation angles, stating that the "5-axis motion sensor" could not detect or compensate for rotation about the Xp axis directly, requiring it to be "derived from the gravitational acceleration detected by the accelerometer." It also noted that accelerometers were unreliable when the device was not static, as they couldn't distinguish gravitational acceleration from other forces.
   • A PHOSITA, aware of these limitations, would have been motivated to incorporate a full three-axis gyroscope to directly measure rotation about all three axes (including Xp) and a three-axis magnetometer to provide an independent, global reference for heading (yaw) that is not subject to gravitational limitations or drift. By 2010, 9-axis inertial measurement units (IMUs) comprising 3-axis accelerometers, 3-axis gyroscopes, and 3-axis magnetometers were known components in the art for achieving robust 3D orientation (Attitude and Heading Reference Systems - AHRS) in applications like robotics, aerospace, and general navigation, where accurate, drift-free 3D orientation was critical.

2. Motivation to employ advanced sensor fusion with quaternions, prediction, and comparison:

   • The patent highlights that Liberty's prior art "cannot accurately or properly calculate or obtain movements, angles and directions of the pointing device while being subject to undesirable interferences... in the dynamic environment" and only provides "relative" movement patterns, leading to errors when a pointer exceeds display boundaries.
   • To overcome these deficiencies and achieve accurate, "absolute" 3D orientation tracking that is robust in dynamic environments and capable of filtering out interferences, a PHOSITA would have been motivated to implement well-known sensor fusion algorithms. Extended Kalman Filters (EKF) and complementary filters, which frequently use quaternions to represent 3D orientation (to avoid issues like gimbal lock), were standard techniques by 2010 for combining noisy data from accelerometers, gyroscopes, and magnetometers. These algorithms inherently involve:
     • Prediction: Estimating the current state (e.g., orientation, angular velocities, predicted accelerations, predicted magnetism) based on the previous state and gyroscope readings. Gyroscopes offer high-frequency, short-term accuracy.
     • Correction/Comparison: Comparing these predicted values with actual measurements from accelerometers (for pitch and roll relative to gravity) and magnetometers (for yaw relative to Earth's magnetic field). Accelerometers and magnetometers provide long-term stability but are susceptible to various disturbances.
     • Update: Fusing the predicted state with the measured discrepancies (innovation) to obtain a refined "updated state" (e.g., an updated quaternion), thereby "eliminating accumulated errors as well as noises over time" and actively "excluding undesirable external interferences."

3. Motivation to include data association:

   • The challenges of external interferences, such as "undesirable axial accelerations caused by undesirable external forces other than a force of gravity" or "undesirable magnetism caused by undesirable electromagnetic fields," were well-known in the field of sensor fusion.
   • A PHOSITA would have been motivated to incorporate known signal processing techniques, such as outlier rejection, adaptive weighting, or data association (as described in the patent), into the sensor fusion algorithm. These techniques are designed to detect and mitigate the impact of unreliable or erroneous sensor measurements, ensuring that the compensation or update steps only utilize trustworthy data. This approach directly addresses the patent's goal of preventing interferences from corrupting the orientation estimate.

Conclusion

It would have been obvious for a PHOSITA, at the priority date of US11698687, to combine the pointing device described in Liberty's prior art with the well-known and generally available 9-axis motion sensor technology and established sensor fusion algorithms (e.g., quaternion-based Kalman or complementary filters, possibly incorporating data association for outlier rejection). The strong motivation for such a combination would be to overcome the acknowledged limitations of Liberty's 5-axis system, particularly its inability to achieve accurate 3D orientation, its susceptibility to dynamic interferences, and its generation of only "relative" movement patterns. By integrating a full 9-axis sensor and implementing a sophisticated sensor fusion algorithm, a PHOSITA would logically expect to achieve a more robust, accurate, and "absolute" 3D motion detection and compensation system, thereby addressing the problems explicitly identified in the background of US11698687.