Reality Computing

Apple getting realer about Reality Computing

Point cloud from a single frame

Apple recently invested in exciting Reality Computing capture technology by buying LinX, an Israeli camera technology company. LinX develops very small multiple-lens cameras that can be incorporated into mobile devices like smartphones and tablets. In addition to capturing 2D still images, the cameras can capture accurate 3D depth information.

The multiple lens are used to calculate tiny discrepancies between images taken at the same time, enabling the creation of detailed 3D depth maps. By capturing both depth information and color, the LinX cameras can create 3D point clouds of objects from a single frame. By combining frames taken from different angles, a point cloud of an entire object or scene can be captured.

That scanned data can then be used in a variety of way (the ‘compute’ part of Reality Computing). For example, you could add or remove objects from a photograph, or use the 3D scanned image of a face in facial recognition software, or calculate the distance between points in a picture.

Once cameras like this become mainstream, everyone will basically have a 3D scanner in their phone…the possibilities are endless!

Click here to read a MacRumors article about LinX’s camera technology.

Reality Computing Power for As-Built Conditions

Reality Computing is helping to transform a century-old decommissioned coal-burning power plant into a new student center for Wisconsin’s Beloit College. The renovated facility will include a conference center, lecture hall, theater, pool, fitness center, gym, and a suspended track winding through the structure. In a nod to the building’s former role, some coal bunkers and steel tubing from the original plant are being preserved and incorporated into the design of the new facility.

The building’s age as well as its numerous catwalks and stairways rendered the existing as-built plans unreliable. So to facilitate their design effort, the project’s design team contacted TURIS Systems, a Wisconsin-based technology consulting group, to capture existing conditions of the building and develop accurate as-built documentation. But the current building is crowded with power generation equipment and many areas of the building and its steel structure are too far out of reach to measure by hand.

 

Reality Computing was used to capture and develop as-built conditions for this century-old power plant, being renovated into a new Beloit College student center. 277 scans were used to capture point cloud data such as this.
Image courtesy of TURIS Systems. 

 

So TURIS turned to Reality Computing in the form of laser scanning (capture) and 3D modeling (compute) to document the as-built conditions. Limited line-of-site issues required TURIS to perform 277 scans to fully capture the building. They used Autodesk ReCap to register the scans and then, with point cloud data as a reference, used Autodesk Revit to model and document the existing building structure, MEP components, power generation equipment, and so on. This accurate as-built information was then turned over to the project architect (Studio Gang) to guide their design effort. The whole process required only three weeks to complete, whereas the design team had budgeted a couple of months to capture that same amount of data—saving time and delivering more accurate information.

For more information about this project, check out this case study video from TURIS. In addition, this AECbytes article, Putting Point Clouds to Work: Then, Now, & Later, profiles the project and also has some very interesting descriptions of how TURIS is using Reality Computing for construction and post-construction applications.

Some call it self-driving. We call it Reality Computing.

It may be another 5 or 10 years before you can buy a self-driving car, but there are already prototypes out there to whet your appetite. And you might think that the venerable Detroit Auto Show would be the main venue for showcasing self-driving cars. Not so. Try the Consumer Electronics Show (CES) in Las Vegas. After this year’s show, Forbes published an article about The Five Most Disruptive Innovations at CES 2015. Guess what was #1?

As you can imagine, self-driving or autonomous vehicles rely heavily on electronics and computing technologies (hence their popularity at CES). A self-driving car is loaded with sensors so it can see the road and environment around it. Extremely powerful computers then process this data and use it to analyze the road and pilot the car—from navigating, steering, and acceleration, to braking and parking.

In other words, the car captures information about the real world around it, computes that data, and from the results creates the driverless transportation experience. Classic example of Reality Computing.

At CES 2015, NVIDIA—a company that a lot of people associate with gaming technology but is also a leading developer of autonomous automotive platforms—demonstrated its new Drive PX and DriveCX computers that enable self-driving cars.

The Drive PX processes data from a range of sensors such as laser, radar, or cameras that capture and monitor the environment around the car. It then uses that live data combined with very sophisticated feature recognition technology it calls ‘deep neural learning’ to dynamically recognize objects around the car including other vehicles, street signs, road works, pedestrians, and so on. The Drive CX is the digital cockpit computer—in charge of speech processing, high-resolution digital instrument display, navigation, and infotainment graphics.

New models of some cars on the market already use Reality Computing to deliver aspects of self-driving like automatic parking, lane departure warning, and adaptive cruise control. But personally, I’m looking forward to the day when I can just get in my car, tell it where I want to go, and then sit back and enjoy the ride!

Reality Computing for Safer Airport Construction

Fort Smith Regional Airport in Arkansas is almost finished with the construction of a three-phase $9 million project that involves upgrading the airport’s primary taxiway, and Reality Computing is help making that construction safer for both travelers and construction crews.

The project entails widening the taxiway 60 feet over a length of about 4,000 feet and includes building a pavement section of 18 inches of concrete over 4 inches of hot mix asphalt, both placed on a compacted subgrade. Arkansas-based Morrison-Shipley is the project’s Engineer of Record and is providing the surveying, engineering design, construction inspection, and administration services for the Airport.

Reality Computing is helping Morrison-Shipley capture existing conditions, and 
assess quality and safety during construction at Fort Smith Regional Airport in Arkansas.

Fort Smith is an active airport, with commercial service as well as operations for private and military aircraft. To help ensure the safety of flight operations, the FAA sets tight limits for construction zones—resulting in a high density of people and construction equipment working in close proximity. This congestion makes it difficult to use traditional surveying methods. So to quickly perform onsite surveys, Morrison-Shipley uses laser-scanning, combined with Autodesk ReCap software to process the resulting point clouds for use in AutoCAD Civil 3D, the firm’s civil engineering design tool. This enables the firm to quickly and safely carry out high-precision data collection at the job site with just one surveyor, greatly reducing risks at the job site.

To capture the existing conditions of the taxiways and surrounding environs, Morrison-Shipley supplemented conventional surveying methods with 3D laser scanning. The firm periodically reused that existing conditions data during the design, avoiding the need for the firm’s survey department to make costly return trips to the site.

In addition, the firm is using laser-scanning technology, ReCap, and Civil 3D to survey in-progress work—documenting subsurface conditions during construction and calculating earthwork quantities. For example, the firm is scanning undercuts of existing subgrade soils and importing these point clouds into Civil 3D to determine exact quantities of excavated material. These measurements are then used by the Airport for the payment of fees to contractors. The firm also uses this reality data to compare as-constructed conditions against its as-designed improvements—verifying the scanned survey data against the finished grades called out on their project drawings.

Click here to read more about how Morrison-Shipley is using Reality Computing to capture and compute this airport construction.