Motion Tracking Overview
July 11, 2013, Motion Tracking Developers’ Conference, San Francisco—Behrooz Abdi from Invensense the efforts and innovation needed to scale and integrate full platforms for our daily lives. The move towards motion tracking differs from simple motion sensing in requirements for accuracy and stability.
The motion sensing market is growing with many more apps becoming available. We are crossing the chasm into an environment with ubiquitous sensing, all functions needing sensors are now available. This market growth is opening the era of awareness in all of our activities. To achieve full motion tracking, devices need cross sensor calibration, data fusion, and the ability to track motion over time.
The history functions are the difference between simple detection and the ability to assist your lifestyle with useful content of your environment. Sensor performance is key. The industry is increasing the number of small and medium scale platforms and is moving sensors into many consumer areas like games, image stabilization, panoramic photos, gesture user interfaces, and in new areas of health and fitness.
The mobility devices have transformed from simple voice instruments to a combination of cloud services and the ability to become a hub or platform for many other functions. Environmental access is becoming a differentiator at all levels and is a large contributor to the growth of big data.
Wearable fitness devices and mobile phones have conducted over 55 M API calls to analytics firms and most of the data are acceleration. Now, the sensors are not just components, but contributors to large data sets, and the machine-to-machine communications will only increase.
The challenges for the hardware are performance, accuracy, drift, power consumption, package size, and cost. The sensors have to provide intelligence and simplicity for the designers and for the users. Integrated sensor platforms will improve time-to-market and simplify the hardware-software integration.
As a fabless company, they had to invent a new manufacturing process that eliminated the cap and glass frit seal approach of other manufacturers. They came up with a bonded wafer approach that also allows for thicker material in the test mass, which increases accuracy and sensitivity. This improvement in accuracy is necessary for indoor tracking, otherwise errors build up. Now, in combination with WiFi and GPS, they can track motion within a building.
For functions like image stabilization, they only need a 2-axis gyro chip. The cameras in the latest smart phones are approaching DSLR quality, and only need better optics to compete. The gyros provide the reference information for optical or electronic image stabilization.
The company has put together a good supply chain with god partners. They have their own test, as this is a key function that requires domain specific expertise. Their partners help to contain the cost structures, and the incremental costs for scaling up production is in the low millions of dollars.
Sensors for tracking are usually the full 9 or 11-axis devices. In addition to the sensors, developers need simple methods to get their data to the app layers, operating system, other software modules, and embedded functions. Their developer’s kit helps with all of these functions, and the sensor fusion software packages the data in a manner that is directly accessible from the OS. They have over 8k developers making new apps for the sensor arrays.
They have OS and app development platforms with integrated fusion tools for the apps and tracking functions. the native OS calls reduce latency and help with power management to provide a better end-to-end user experience. The integrated solutions use second-generation motion software to produce context-aware navigation and smooth imaging applications.
