This Small Business Innovation Research (SBIR) Phase I project proposes to dramatically advance 3D position input and motion sensing systems with a MEMS-based optical 3D tracking and 3D position measurements system. This sensing system would enable real time interaction with computers, robots, and other machinery in ways that are more intuitive, precise and natural compared to existing technologies. Using scanning MEMS mirrors, lasers, optics and optical sensors, this tracking system can quickly and precisely determine location, direction, and rate of movement. Real time 3D tracking has many potential applications. The proposed optical sensor will provide X,Y, and Z position information of a small target object down to sub-millimeter precision in large spaces such as a room and at update rates below 0.1ms. These might include controllers for video games, PC interactions, industrial robot controls, crane operations, etc. Having controllers that can follow our hands or fingers and interpret our intentions would make human-machine interaction a more enjoyable, ergonomic, and precise experience. Mirrorcle Technologies has been a technical leader in the area of MEMS mirrors. However, building such a 3D tracking system requires significant innovation in optical design, closed-loop control, embedded computing, and packaging.
The broader impact/commercial potential of this project is to enable a superior, more economical, and more accurate means of interacting with machines by humans. Human machine interaction is a necessity in our modern world. The number of machines we interact with will continue to increase, and interactions will be increasingly complex. However, human-machine interaction is often not pleasant and intuitive. Ergonomic injuries caused by keyboard, mouse, and poor posture are increasingly common. Developing capabilities for more intuitive interaction will serve a great need for our increasingly electronic world. Adding more knobs and buttons is not the right solution. We believe that input devices based on 3D tracking can be developed into intuitive and natural ways for people to interact with augmented and virtual reality environments. It may enhance our ability to work with objects much bigger or smaller than our range of manual dexterity. Ultimately, it can help us navigate an ever more complex computerized world.
Mirrorcle Technologies is developing a tracking and position measurement technology based on a scanning MEMS micro-mirror capable of searching and tracking the position of a retro reflector. The device is a complex system with many components (lasers, optics, electronics, and MEMS device,) that must ultimately work together with proper synchronization, and space and power utilization. The key component of the system is MirrorcleTech’s proprietary gimbal-less MEMS mirror which achieves high-speed point-to-point optical scanning in two-axes. These devices consume milliwatts of power and very small volumes while achieving high performance laser-beam scanning required for a variety of laser-based imaging and metrology applications. Obtaining real-time 3D coordinates of a moving object has many applications such as gaming, robotics and human-computer interaction applications, and industrial applications such as metrology and factory automation. Various technologies have been investigated for and used in these applications, including sensing via wire-interfaces, ultrasound, and laser interferometry. However a simple and low-cost solution that can provide enough precision and flexibility is not yet available. In Phase I we proposed two different beam-steering based schemes or techniques to track an object inside a conic volume. The two versions were both successfully designed and implemented, as we reported in the final Phase I report. MTI is happy to report significant findings and progress in its research using MEMS micromirror-based optical units (recently named "MEMSEyes") as position and orientation sensors. Specifically: 1) Successfully completed designs and validated the two proposed methods of optically tracking objects. Used optical CAD and analysis to design and validate the trade-offs and limitations of the two design types we had proposed for phase I. 2) Both design types were built and successfully demonstrated. Verified that both designs function as expected and provide sufficient signal and information to allow the system to point the MEMS mirror in the direction of the tracked object even as that object is moving. 3) Completed comparison of the two methods and selected the retro-reflector tracking version over the LED tracking version. As part of Phase IB effort we report the following significant findings and progress: 4) The package design of MEMS mirror was accomplished and a large number of test samples and packages were prepared for trials with several packaging vendors. 5) The opto-mechanical design of our tracking sensor (arrangement of micro-mirror and associated optics and photo sensor) using the new MEMSEye design was completed. The design saves space and greatly improves alignment and optical efficiency. 6) Completed a prototype with two MEMSEyes to demonstrate our overall 3D position measurement technique by triangulation, taking in position data and calculating a line vector from an origin to the target object. 7) Conducted successful experiments of orientation measurement of a remote object: tracking of two targets (corner cube retro-reflectors) located on a rod in 3D space, thereby allowing us to calculate the rod’s orientation (azimuth and elevation of the rod.) During this visit the precision of the system was found to be within +/- 0.1° in determining the rod’s orientation. 8) Presented an Invited talk at the SPIE Photonics West 2011 in San Francisco, CA and published the paper entitled: ""MEMSEye" for Optical 3D Position and Orientation Measurement". 1.2 Position Measurement - Results from Phase I Multiple prototype arrangements were tested. LED tracking tests required the use of significantly larger mirror diameters in order to capture enough light from the LED once the LED was a significant (~1m) distance away from the sensor. This matched expectations based on simulation results. With a 3.2mm mirror diameter and an aperture to block undesired reflections to the photo sensor, tracking and position measurement was demonstrated up to about 1.25m distance. In the closer range where tracking was successful, the LED movements could not be made very fast because the system’s loop bandwidth was reduced to a much slower speed to accommodate the large mirror diameter and its low resonant frequency. Retroreflector tracking prototypes performed at greater distances, wide angles, and due to the use of a small mirror (1mm diameter,) significantly greater speeds of target motion were trackable. Robust tracking of both corner cube retro-reflector (CCR) targets, as well as retro-reflective tape targets was demonstrated. The MEMSEye system was able to track and follow the individual position of the retro-reflective tape placed on the tip of a pencil, or on the edge of a cell phone, in a wide-angle cone of approx. 45°. After some preliminary system calibrations by approximating the angle that each MEMS mirror points to at a given voltage, the XYZ determination algorithm was tested. With preliminary calibration distances are found to be accurate within a few mm in all 3 directions, in a large volume of over 1m3. Precision and repeatability are better than 1mm in distance (Z) and better than 0.1mm in X and Y.
