August 2014 focused on evaluating new sensor options while engaging with the technical community through meetings and competition entries.
Structure IO Sensor Evaluation
The Structure Sensor finally arrived, bringing high expectations based on its marketing. Reality proved disappointing. The sensor wasn’t suitable for close proximity scanning—performing worse than the Creative and SoftKinetic sensors already in use. Even its 3D object scanning capabilities, a primary selling point, fell short of expectations. This evaluation reinforced that finding the right depth sensing hardware remained one of the project’s fundamental challenges.
Microsoft Research Investigation
An interesting connection emerged through industry contacts: a reference to Microsoft Distinguished Scientist Steven Bathiche, known for work in new UI areas. His brief viewing of demonstration videos (31 seconds before stopping) and subsequent email indicating he’d “seen this sort of thing” before prompted deeper research.
Investigation into Bathiche’s patent portfolio initially found nothing overlapping with Visual Touchscreens concepts. However, broader searching revealed Microsoft’s RetroDepth project—work in a similar technical area but with different applications. Further research into related researchers like David Kim found multiple papers in the general space of depth-sensing interaction, though none that specifically obviated the Visual Touch concepts.
This research served dual purposes: understanding competitive landscape and assessing patent defensibility. The reality that large companies like Microsoft were exploring similar technical territories was simultaneously validating and concerning.
Hardware Experimentation
Continued sensor experimentation included working with the Duo3D stereo camera. Disassembling the lenses revealed how focal length affected image quality—unscrewing them completely resulted in gray output, while finding the right adjustment restored focus.
Research expanded to CCD sensor options, particularly medical industry flat-panel sensors used for X-rays. These large-format sensors resembled what would be ideal for visual touch systems like Pixelsense, though cost and availability made them impractical for prototyping.
Fresnel lens arrays emerged as another possibility worth investigating. Each hardware research path opened new avenues while revealing practical constraints around cost, availability, and integration complexity.
Competition Entry
An idea was submitted to the Intel RealSense challenge, continuing efforts to gain visibility through competitions and challenges. These programs offered structured ways to present the technology to relevant industry audiences, even when winning wasn’t guaranteed.
Imaging Industry Research
Deep dives into imaging sensor companies provided ongoing education. Following blogs, Twitter feeds, and Google+ pages from multiple imaging sensor companies became part of staying current with rapidly evolving hardware capabilities. The image-sensors-world blog proved particularly valuable for tracking industry developments.
Questions to hardware vendors like Duo3D about connecting multiple devices concurrently reflected thinking about system architecture beyond single-sensor prototypes. Future product versions might require multiple depth cameras working together.
The Curtin Ignition Decision
Late in the month, preparation began for the upcoming Curtin Ignition startup accelerator program. The decision to participate represented commitment to structured business development alongside continued technical work. September would bring intensive engagement with this program, temporarily shifting focus from pure R&D to customer discovery, business model development, and pitch refinement.
The month demonstrated the multifaceted nature of moving innovation toward market: technical evaluation of hardware continued, but community engagement, competitive research, and accelerator participation became equally important workstreams. Building a successful product required capabilities beyond engineering.