Andrea Danti - Fotolia
Reality computing will have a profound effect on the development of 3-D technology. The emergence of reality computing is likely to bring new opportunities, as well as integration and architectural challenges, around connecting disparate sensor systems, managing terabytes of data, and integrating across differing applications and associated data sets.
"Reality computing involves a variety of tools and technologies that don't play well together at the moment," said Chet Roan, business development manager at Epic Scan, a reality computing consultancy at the Real 2015 conference in San Francisco. Enterprises often need to lean on cross-disciplinary experts able to reliably and practically get data from specialized sensors in order to improve business processes, he said.
Making sense of reality computing
The Real 2015 conference was an experiment on Autodesk's part to see what could happen when you bring together experts across disciplines as varied as engineering, design, architecture, fashion and art, said Alonzo Addison, Real 2015 co-chair and professor at OCAD University in Toronto. Each of these disciplines has been focused on working with point solutions to improve the use of 3-D technology. But Addison feels an opportunity exists for cross-disciplinary sharing to help drive the field of reality computing forward as a whole.
Mark Yahiromanaging director for 3-D perceptual computing at Intel
The main thrust of reality computing lies in figuring out how to capture 3-D data from the world and process this data into useful models. This opens the door for applications in engineering, product development and consumer experiences. This is no easy task because each set of tools gathers, processes and stores data in different ways. Likewise, the tools that use this data, such as CAD systems, building management systems, 3-D printing and virtual reality rendering work with different formats.
For capturing data, the gold standard is the use of LIDAR to quickly take 3-D measurements of the built environment as point cloud data that can require gigabytes of storage. The sensors themselves range in price from $20,000 for a basic tool to $750,000 for a truck-mounted rig that can automatically measure roads at high speed. Keith Hay, assistant surveyor at R.E.Y Engineers, said that using truck-mounted LIDAR gear from Riegl has allowed a team of two to complete a road survey in two days. Without this equipment a survey would take a much larger team three months to complete.
New tools required
After cloud data has been captured, it must be transformed into a 3-D representation. Traditionally, this has been a time-consuming process. New software tools such as Autodesk's ReCap 2.0 promise to streamline this process.
Other methods of data capture promise to gather the raw data required for reality computing much more cheaply using techniques like structured light processing, depth sensors and photogrammetry, which is the process of stitching multiple photos together to derive accurate measurements. The midlevel tools range in price from $5,000 to $50,000, and can leverage a new generation of cloud services provided by companies like Artec Group, Matterport and Floored to convert this data into useful models.
For example, Artec Group has developed the Shapify service, which captures a 3-D model of a person in 12 seconds. The model can be converted into a 3-D printed statue. Both Matterport and Floored have developed cloud services for the real-estate industry that can use 3-D sensor data or floor plans for generating commercial and residential virtual reality views. These can speed real estate transactions for commercial properties and increase the amount of time consumers spend looking at a property online, said David Eisenberg, founder and CEO of Floored.
Bringing 3-D to the masses
A variety of companies, including Intel RealSense, Hewlett Packard Sprout and Occipital Structure, are developing consumer-priced 3-D sensors that promise to bring reality computing to the masses for a few hundred dollars today, and for as little as a few dollars in the near future. Intel is working with leading computer and tablet manufacturers including Dell and Lenovo to embed the RealSense technology into new mobile devices and industrial equipment. For example, Nordstrom is using RealSense in a pilot program to more accurately capture shoe sizes.
"Our challenge was to make this available to everyone in the world," said Mark Yahiro, managing director of 3-D perceptual computing at Intel. Intel has released a free SDK with reality computing capabilities for developers to build new applications. Yahiro believes this will open the door for a variety of new business and consumer use cases that were expensive before.
Autodesk released a new software package and cloud service called Memento that allows anyone to stitch together a 3-D model of a space or an environment from a collection of photos. Memento can output 3-D virtual views and produce printable 3-D objects. This is already being used by the Smithsonian to make more of its collection available, allow people to play with fossils and to explore changes in coral reef systems.
Metaio is leveraging the Memento platform to create new applications for overlaying 3-D representations of objects and furniture onto the physical world. They are working with IKEA to create virtual furniture that can be placed and moved around a living room. These can be affixed to specially printed posters, for example, so that consumers can experiment with different layouts virtually before buying a new couch.
The software is in its early stages, said Epic Scan's Roan. Even though Memento is a vast improvement on the point solutions that now exist, companies will need technical guidance to create practical reality computing solutions.
IT Challenge: Using 3-D printing in manufacturing
Discover 10 recent innovations in 3-D printing