Image Sensors at 2016 EI Symposium

...
2016 IS&T International Symposium on Electronic Imaging to be held on Feb. 14–18 in San Francisco, CA, publishes its preliminary Program. There are many image sensor related short courses and papers:

EI13: Introduction to CMOS Image Sensor Technology

Instructor: Arnaud Darmont, APHESA

A time-of-flight CMOS range image sensor using 4-tap output pixels with lateral-electric-field control,

Taichi Kasugai, Sang-Man Han, Hanh Trang, Taishi Takasawa, Satoshi Aoyama, Keita Yasutomi, Keiichiro Kagawa, and Shoji Kawahito;
Shizuoka Univ. and Brookman Technology (Japan)

Design, implementation and evaluation of a TOF range image sensor using multi-tap lock-in pixels with cascaded charge draining and modulating gates,

Trang Nguyen, Taichi Kasugai, Keigo Isobe, Sang-Man Han, Taishi Takasawa, De XIng Lioe, Keita Yasutomi, Keiichiro Kagawa, and Shoji Kawahito;
Shizuoka Univ. and Brookman Technology (Japan)

A high dynamic range linear vision sensor with event asynchronous and frame-based synchronous operation,
Juan A. Leñero-Bardallo, Ricardo Carmona-Galán, and Angel Rodríguez-Vázquez,
Universidad de Sevilla (Spain)

A dual-core highly programmable 120dB image sensor,

Benoit Dupont,
Pyxalis (France)

Analog current mode implementation of global and local tone mapping algorithm for wide dynamic range image display,
Peng Chen, Kartikeya Murari, and Orly Yadid-Pecht,
Univ. of Calgary (Canada)

High dynamic range challenges
Short presentation by Arnaud Darmont, APHESA SPRL (Belgium)

Image sensor with organic photoconductive films by stacking the red/green and blue components,
Tomomi Takagi, Toshikatu Sakai, Kazunori Miyakawa, and Mamoru Furuta;
NHK Science & Technology Research Laboratories and Kochi University of Technology (Japan)

High-sensitivity CMOS image sensor overlaid with Ga2O3/CIGS heterojunction photodiode,
Kazunori Miyakawa, Shigeyuki Imura, Hiroshi Ohtake, Misao Kubota, Kenji Kikuchi, Tokio Nakada, Toru Okino, Yutaka Hirose, Yoshihisa Kato, and Nobukazu Teranishi;
NHK Science and Technology Research Laboratories, NHK Sapporo Station, Tokyo University of Science, Panasonic Corporation, University of Hyogo, and Shizuoka University (Japan)

Sub-micron pixel CMOS image sensor with new color filter patterns,
Biay-Cheng Hseih, Sergio Goma, Hasib Siddiqui, Kalin Atanassov, Jiafu Luo, RJ Lin, Hy Cheng, Kuoyu Chou, JJ Sze, and Calvin Chao;
Qualcomm Technologies Inc. (United States) and TSMC (Taiwan)

A CMOS image sensor with variable frame rate for low-power operation,
Byoung-Soo Choi, Sung-Hyun Jo, Myunghan Bae, Sang-Hwan Kim, and Jang-Kyoo Shin, Kyungpook National University (South Korea)

ADC techniques for optimized conversion time in CMOS image sensors,
Cedric Pastorelli and Pascal Mellot; ANRT and STMicroelectronics (France)

Miniature lensless computational infrared imager,
Evan Erickson, Mark Kellam, Patrick Gill, James Tringali, and David Stork,
Rambus (United States)

Focal-plane scale space generation with a 6T pixel architecture,
Fernanda Oliveira, José Gabriel Gomes, Ricardo Carmona-Galán, Jorge Fernández-Berni, and Angel Rodríguez-Vázquez;
Universidade Federal do Rio de Janeiro (Brazil) and Instituto de Microelectrónica de Sevilla (Spain)

Development of an 8K full-resolution single-chip image acquisition system,
Tomohiro Nakamura, Ryohei Funatsu, Takahiro Yamasaki, Kazuya Kitamura, and Hiroshi Shimamoto,
Japan Broadcasting Corporation (NHK) (Japan)

A 1.12-um pixel CMOS image sensor survey,
Clemenz Portmann, Lele Wang, Guofeng Liu, Ousmane Diop, and Boyd Fowler,
Google Inc (United States)

A comparative noise analysis and measurement for n-type and p-type pixels with CMS technique,
Xiaoliang Ge, Bastien Mamdy, and Albert Theuwissen;
Technische Univ. Delft (Netherlands), STMicroelectronics, Universite Claude Bernard Lyon 1 (France), and Harvest Imaging (Belgium)

Increases in hot pixel development rates for small digital pixel sizes,
Glenn Chapman, Rahul Thomas, Rohan Thomas, Klinsmann Meneses, Tony Yang, Israel Koren, and Zahava Koren;
Simon Fraser Univ. (Canada) and Univ. of Massachusetts Amherst (United States)

Correlation of photo-response blooming metrics with image quality in CMOS image sensors,
Pulla Reddy Ailuri, Orit Skorka, Ning Li, Radu Ispasoiu, and Vladi Koborov;
ON Semiconductor (United States)
2:52 AM

Light Camera Demo

...
Mashable got a chance to see Light Co's L16 52MP array camera prototype at CES 2016. Few quotes from Mashable impressions:

"On the prototype, the photo stitching took a little while to work and froze. In the end, I didn't get to see how fast it was. When I asked Dr. Rajiv Laroia, Light's co-founder and Chief Technology Officer, how long it will take to generate a 52-megapixel image on the final product, he told me they're shooting for under a minute.

That's a long time to wait for a complete image. The Light team is going to try to make the processing as fast and instantaneous as possible, but the company's not promising anything faster than under a minute right now.
"

"I have to admit, the sample image taken by the L16 looked pretty good with lots of details when zoomed in, but it also looked like it had a lot of image noise."

12:52 AM

Lenovo to Make Smartphone with Google Tango 3D Camera

...
VentureBeat reports that Lenovo and Google announce a partnership to create a smartphone with 3D camera-based Project Tango. Jeff Meredith, Lenovo VP, said the goal was to make a mainstream device scheduled to appear on the market on summer 2016.

12:11 AM

Socionext Shipping Dual Camera Image Processor

...
PRNewswire: Socionext (Fujitsu + Panasonic Semi) introduces “M-12MO” (MBG967) Milbeaut Image Processor. The MBG967, which will be available in volume shipments starting in January, is mainly targeted at smartphones and other mobile applications. It supports dual camera, the latest trend in mobile applications, along with functionalities such as low light shot and depth map generation. The expansion of dual camera capabilities in the mobile camera market has been highly anticipated because dual cameras enable new functionalities previously considered difficult with mobile cameras. These include low light shot, which integrates images from color and monochrome sensors, and the generation of depth maps, which can create background blur comparable to that of SLR cameras.

Main features of the MBG967 include:

Low light shot by dual camera: By integrating the images from color and monochrome image sensors, the MBG967 enables high-sensitivity, low-noise pictures:


High-speed, high-accuracy auto focus supports high speed “Phase Detect AF”, in addition to conventional “Contrast AF”. The MBG967 also supports “Laser AF” which has an advantage in the low light conditions. Its “Super Hybrid AF” utilizes these three AF methods in combination, always allowing faster and more accurate AF in varying conditions:

1:21 PM

Intel Unveils R200 and ZR300 RealSense 3D Cameras

...
Intel announces R200 RealSense 3D camera, said to be the company's first long-range depth camera for 2 in 1s and tablets. The new camera is aimed to:
  • 3D scanning: Scan people and objects in 3D to share on social media or print on a 3D printer.
  • Immersive Gaming: Scan oneself into a game and be the character in top rated games
  • Enhanced Photography/Video: Create live video with depth enabled special effects, remove/change backgrounds or enhance the focus and color of photographs on the fly.
  • Immersive Shopping: Capturing body shape and measurements as depth data that is transformed into a digital model enabling people to virtually try on clothes.
The RealSense R200 camera is capable of capturing VGA-resolution depth information at 60 fps. The camera uses dual-infrared imagers to calculate depth using stereoscopic techniques. By leveraging IR technology, the camera provides reliable depth information even in darker areas and shadows as well as when capturing flat or texture-less surfaces. The operating range for the Intel RealSense Camera R200 is between 0.5 meters and 3.5 meters, in indoor situations. The RGB sensor is 1080p resolution at 30 fps.

A number of OEM featurs RealSense R200, including the HP Spectre x2, Lenovo Ideapad Miix 700, Acer Aspire Switch 12 S, NEC LaVie Hybrid Zero11 and Panasonic. The Intel RealSense Camera R200 is supported on all Windows 10 systems that run on 6th Generation Intel Core Processors.

RealSense ZR300 camera is an integrated unit within the new RealSense Smartphone Developer Kit. The Intel RealSense Camera ZR300 provides high-quality and high-density depth data at VGA-resolution of 60 fps. The ZR300 supports Google Project Tango spec for feature tracking and synchronization via time stamping between sensors.

Source: The Inquirer
1:05 PM

NVIDIA Presents Deep Learning Automotive Imaging Platform

...
NVIDIA launches NVIDIA DRIVE PX 2, said to be the world’s most powerful engine for in-vehicle artificial intelligence. DRIVE PX 2 can process the inputs of 12 video cameras, plus lidar, radar and ultrasonic sensors. It fuses them to accurately detect objects, identify them, determine where the car is relative to the world around it, and then calculate its optimal path for safe travel.

The company's Youtube promo video presents the new processor:

1:28 PM

14nm Ambarella Camera SoC Consumes 2W in 4K 60fps Mode

...
BusinessWire: Ambarella introduces the H2 and H12 camera SoCs for sports and flying cameras. 14nm process-based H2 targets high-end camera models with 4K Ultra HD H.265/HEVC video at 60 fps and 4K AVC video at 120 fps and includes 10-bit HDR video processing. 28nm process-based H12 targets mainstream cameras and offers 4K Ultra HD HEVC video at 30 fps.

“With the introduction of H2 and H12 we now provide a complete portfolio of 4K Ultra HD HEVC solutions for sports and flying cameras,” said Fermi Wang, President and CEO of Ambarella.

1:14 PM

Himax WLO Adopted in 3D Structured Light Camera

...
Himax announces that its Wafer Level Optics (“WLO”) laser diode collimator with integrated Diffractive Optical Element (“DOE”) has been integrated into laser projectors for next-generation structured light camera. Himax's WLO systemhas a height of less than two millimeters. The WLO component is then stacked on top of a laser diode to reduce the overall height of a coded laser projector assembly to five millimeters.

Jordan Wu, President and CEO of Himax Technologies says "We are currently collaborating with several major OEMs' product developments using our WLO as our expertise in WLO design and manufacturing enables significant size and cost reduction of coded laser projectors. For example, in an active sensing 3D camera projector, our technology can reduce the size of the incumbent laser projector module by a factor of 9, actually making it smaller than conventional camera modules. This breakthrough allows our WLO collimator to be easily integrated into next-generation smartphones, tablets, automobiles, wearable devices, IoT applications, consumer electronics accessories and several other products to enable new applications in the consumer, medical, and industrial marketplaces."

The WLO laser collimator and DOE will be manufactured by Himax’s Wafer Optics production facility in Taiwan. The first production run for 3D camera applications is scheduled for delivery and sampling by Himax's partners and select
customers in Q1 2016.

Update: GlobeNewsWire: Himax reports "higher-than-expected engineering fees from AR/VR project engagements with both current and new customers."
1:03 PM

Arctic Sea Ice At Record Low

...
Arctic sea ice extent on January 4, 2016, was at a record low for the time of the year, as illustrated by the image below.


Arctic sea ice will typically reach its maximum extent in March. In 2015, sea ice extent was very low in March (see blue line in above image), and the outlook for this year is even more grim, as oceans get warmer and El Niño is still gaining in strength.

Below is a comparison of sea ice thickness (in m) on January 4th for the years 2012, 2015 and 2016.


Below is an update showing Arctic sea ice extent as on January 6, 2016, at the bottom left corner, marked with the red dot.



The situation is dire and calls for comprehensive and effective action as described at the Climate Plan.



Arctic sea ice extent on January 4, 2016, was at a record low for the time of the year. Arctic Sea Ice At Record Lowhttp://arctic-news.blogspot.com/2016/01/arctic-sea-ice-at-record-low.html
Posted by Sam Carana on Tuesday, January 5, 2016

Sandisk VP Predicts 80MP Mobile Cameras This Year

...
EETimes: Christopher Bergey, VP & GM of Mobile and Connected Solutions, SanDisk, makes mobile imaging predictions for 2016:

"As we move into 2016, we will see major strides made towards a radically new smartphone camera market. Camera makers are pushing the boundaries of technology and exploring new areas, such as 3D cameras, massive megapixels (80MB), cameras that can take 360 degree panoramic images and video and cameras that can shoot 1,000 frames a second. 4K Ultra HD for mobile is another move to watch in 2016. Not just for video streaming, we’ll see users take advantage of this extreme resolution in new ways as they shoot and create their own content on their smartphones."
1:32 PM

ST Presents 2nd Gen ToF Ranger Working at 940nm Wavelength

...
GlobeNewsWire: STMicroelectronics releases its second-generation FlightSense sensor for mobile devices. The new VL53L0 is faster, working over longer distances, and more accurate.

With its form factor of 4.4 x 2.4 x 1mm, the 6-pin packaged VL53L0 is said to be the smallest ToF module in the world, and the first to integrate a 940nm VCSEL light source, a SPAD photon detector, and an advanced microcontroller to manage the complete ranging function. Being the market's first module to use light emitted at 940nm, coupled to leading-edge infrared filters, the VL53L0 is said to deliver best-in-class ambient light immunity and is now invisible to the human eye.

"ST technology advancements in Time-of-Flight ranging sensors are enhancing the experience for millions of consumers, revolutionizing the way they take pictures and videos with their smartphones and tablets," said Eric Aussedat, GM of ST's Imaging Division. "ST introduced the first fully-integrated Time-of-Flight ranging sensor to the market in 2014, which was then successfully adopted by several leading OEMs for the laser-assisted autofocus function. Today, with the VL53L0, our next generation, ST is redefining the benchmark in ranging performance and creating the opportunity to develop new applications in robotics and the IoT."

The VL53L0 is able to perform a full measurement operation in one image frame, typically less than 30ms, at distances beyond 2m. With such performance levels the camera system can achieve instant focus in both video and burst modes, even in low-light or low-contrast scenes.

The VL53L0 is in production and available now, priced from $1.75 at the minimum order quantity of 5,000 units.

12:48 PM

Huawei Invested $98M in Smartphone Image Processing

...
BusinessWire: Huawei invested $98M over three years to develop its first proprietary image sensor processor for faster focusing, higher clarity, and more accurate color shading. A leading research team was assembled in France to increase the processing speed of the camera, resulting in increasing the process bandwidth by four times with 14b precision. (Possibly, Huawei talks here about ex-TI OMAP team in Nice, France, that TI has laid off 3 years ago, and Huawei hired - ISW)
11:52 AM

Gestigon Ports its Firmware to PMD and Inuitive Hardware

...
gestigon and pmdtechnologies announce a collaboration that combines Samsung’s GearVR, pmd’s CamBoard pico flexx depth sensor and gestigon’s Carnival AR/VR Interaction Suite to showcase how existing VR devices can be augmented with touchless interaction.

"pmd is proud to contribute our advanced depth sensing technology to gestigon’s effort,“ states Bernd Buxbaum, CEO of pmdtechnologies. “We are excited to bring a new VR user experience to Samsung’s GearVR headset,” confirms Moritz von Grotthuss, CEO of gestigon.

Current ways to interact with Gear VR applications are extremely limited, involving turns and nods of the head to indicate menu choices. gestigon’s Carnival SDK enables a more natural interaction by visualizing the user’s hands within the application and using gesture recognition to choose from multiple choice menus or interact with virtual objects. The Carnival SDK requires depth information generated by pmd’s pico flexx sensor, which is mounted in the front of the Gear VR headset and connected to the smartphone’s USB port:


Inuitive, a developer of 3D computer vision and image processors, and gestigon announce a collaboration to bring gesture recognition to embedded virtual reality platforms.

“Using today’s head-mounted VR displays, my hands are either not visible, or the tracking is so slow and inaccurate that the hands feel more like a robot’s and not my own,” says Moritz v. Grotthuss, CEO of gestigon.

Second generation head-mounted-devices will include front-facing 3D sensors to improve realism, but component cost and power consumption are key concerns. Bringing together the Inuitive NU3000 multi-core imaging processor and gestigon gesture recognition algorithms, the collaboration between the two companies aims to address these concerns.

“Our unique technology and architecture uses input from standard, low-cost cameras to efficiently generate depth maps. Now, through our collaboration with gestigon, we can offer a complete one-stop solution to our customers, shortening the development cycle,” said Shlomo Gadot, CEO and Co-Founder of Inuitive.

Inuitive’s NU3000 processor incorporates two CEVA MM3101 high-performance, low-power imaging and computer vision vector DSP cores. In addition, it integrates a dedicated hardware accelerator capable of extracting real-time depth maps from stereo vision input. The gestigon gesture recognition algorithms, based on its Carnival AR/VR Interaction Suite, are customized and optimized to run directly on this processor to provide fingertip and hand tracking, as well as gesture recognition.

11:18 AM

Heptagon Introduces TARO 3DRanger, a Smart ToF Camera

...
BusinessWire: Heptagon introduces TARO, a ToF 3DRanger Smart Camera for fast, 3D depth sensing applications up to several meters. TARO is said to be the industry-first complete solution that combines a 3D ToF Camera with on-board algorithms and software. TARO improves upon Heptagon’s successful SwissRanger line of industrial ToF cameras, with a match-box sized footprint, on-board image processing for fast, easy-to-use 3D depth information, and the ability to download new algorithms into the TARO hardware. This makes TARO a powerful and versatile platform for retail analytics like people counting, occupancy detection, and intruder detection.

11:01 AM

Omnivision CameraCubeChip Features in Eye Tracking Device

...
PRNewswire: OmniVision's global shutter CameraCubeChip will be integrated into SMI eye tracking platform. Using OmniVision's 3 um OmniPixel3-GS global shutter pixel, the OVM6211 features high sensitivity in NIR, minimizing the required power budget of the NIR light source. The OVM6211 captures full resolution 400x400 pixel video at 120 fps. CameraCubeChip technology enables the OVM6211's extremely compact form factor of 3.23 x 3.23 x 3.92 mm for space constrained devices.

A year-old Youtube video presents SMI and Omnivision partnership:

10:51 AM

To Blog Followers Not Using Google Accounts: Change in Google Policy

...

This notice is for all followers of this blog who use Twitter, Yahoo, Orkut or other "OpenID" providers to follow this blog.  On Jan. 11, Google is going to require you to have a Google account in order to follow this blog.  If you do not have a Google account and wish to continue following this blog, you need to take action before Jan. 11.  Here is the full text of the information I received on Jan. 4:

An update on Google Friend Connect

Posted: December 21, 2015
In 2011, we announced the retirement of Google Friend Connect for all non-Blogger sites. We made an exception for Blogger to give readers an easy way to follow blogs using a variety of accounts. Yet over time, we’ve seen that most people sign into Friend Connect with a Google Account. So, in an effort to streamline, in the next few weeks we’ll be making some changes that will eventually require readers to have a Google Account to sign into Friend Connect and follow blogs.

As part of this plan, starting the week of January 11, we’ll remove the ability for people with Twitter, Yahoo, Orkut or other OpenId providers to sign in to Google Friend Connect and follow blogs. At the same time, we’ll remove non-Google Account profiles so you may see a decrease in your blog follower count.

We encourage you to tell affected readers (perhaps via a blog post), that if they use a non-Google Account to follow your blog, they need to sign up for a Google Account, and re-follow your blog. With a Google Account, they’ll get blogs added to their Reading List, making it easier for them to see the latest posts and activity of the blogs they follow.

We know how important followers are to all bloggers, but we believe this change will improve the experience for both you and your readers.

Posted by Michael Goddard, Software Engineer
4:02 AM

Sandwich Panels and the Dubai Hotel Fire

...

If you were watching TV on New Year's Eve, amid all the spectacular fireworks displays in cities around the world you might have also seen an unplanned spectacle:  the blaze climbing up one side of the 63-story Address Hotel in Dubai, the largest city of the United Arab Emirates (UAE).  While the Address Hotel is not the tallest skyscraper in the world (that honor goes to the Burj Khalifa, also in Dubai), it's tall enough to attract global attention as it was enveloped in flames during the night.  Amazingly, no fatalities were reported, although numerous people suffered smoke inhalation or minor injuries while the hotel was being evacuated.  The main reason for the absence of serious casualties was that the fire was confined almost entirely to the "sandwich panels" or cladding on the outside of the building.  Why they could catch fire—and why entire buildings are covered with flammable material in the first place—are topics worth pursuing.

As most people know, modern skyscrapers depend on a hidden steel skeleton for mechanical strength, not on the exterior surfaces, which can be chosen for properties other than their ability to support the building.  At first, high-rise architects stuck to the traditional stone, concrete, and brick for facades, but in the 1950s, they began to experiment with lighter-weight and cheaper materials, such as glass and aluminum. Properly handled and mounted, aluminum makes a fine, long-lasting sheathing material, and so does glass.  Then a couple of decades ago, someone had the idea of sandwiching a few millimeters of plastic—polyethylene or some other heat-softening—between two thin  foil-like claddings of aluminum, making something cheaper and lighter but just as good-looking as solid aluminum.  Thus the sandwich panel was born.

Now, most heat-softening (thermoplastic) plastics can burn very easily, and some attempts were made to introduce fire-retardant materials into the plastic core of sandwich panels to make them fire-resistant.  Apparently, these attempts did not convince U. S. building-code authorities that the new sandwich panels were safe enough to use in high-rises.  A comment on an architect's chatroom I found indicates that these types of panels are prohibited in the U. S. for use on buildings taller than about four stories.  But other countries either had no such laws, or came to realize the potential for disaster too late.

What can happen is this.  If you have a whole building encased in this stuff, and one panel near the bottom happens to catch fire somehow (fireworks seem to be a popular way to do this), you are in big trouble.  Aluminum has a low melting point and melts away from the plastic cladding as soon as the flame reaches it, exposing more plastic to air and letting the fire feed on itself.  Hot air and flames travel upward to the next panel and so on, and in the case of a 63-story building, there's plenty of upward to travel through.  This is what happened, apparently, not only to the Address Hotel, but to several other similarly-clad high-rises in Dubai and elsewhere in the last few years.  The architect-chatroom website where this problem was discussed has numerous pictures of burned-over building exteriors in Dubai, China, and elsewhere—all fires in which sandwich panels played a critical role.

Fortunately, the fires that these panels support tend to stick to the outside, and most of the time, people inside the buildings have time to evacuate before anyone gets killed.  But nobody wants to leave a building under duress while dodging falling pieces of burning plastic and metal on your way out.  And it's very costly to clean up the resulting mess and re-cover the structure with something that won't burn as easily next time.

Both Australia (where such a fire happened in Melbourne in 2014) and the UAE have changed their building codes to require sandwich panels to pass certain fire-retardant tests.  There are two problems with this, however.  One, it's not clear exactly how fire-retardant a panel has to be in order to resist spreading a fire on a tall building.  The only sure way to know is to build such a building and try to set fire to it, and this experiment is beyond the resources of most building-code-writing organizations.  Second, such codes generally apply only to new construction, and are not retroactive.  So anyone who's already built a skyscraper with flammable cladding doesn't have to take the cladding down and replace it with something better.  That is, until it catches fire.  Judging by the fact that the Address Hotel fire was the third such conflagration in Dubai in three years, it may be only a matter of time until the others light up too.

Modeling how fires start and spread is still an inexact science, and it is understandable that pressures from the building industry allowed dangerous sandwich panels to be installed in many places around the world, despite the hazards involved.  But it takes only one or two fires like this to demonstrate that there's a serious problem.  The almost universal tradition of not making building codes retroactive makes sense, because taking stuff out of an existing building to replace it can be more expensive than the original building cost.  Better in that case simply to condemn the thing and tear it down, but that's an extreme measure too. 

So what's the best that can be done in the present situation?  There may be some lower-cost ways to reduce the chances that a fire in existing sandwich panels will spread, possibly by installing some kind of fire-break strip at selected heights.  But that would be pretty speculative and might not work.  Another proposal has been to install fire sprinklers on balconies near sandwich panels, because many of the buildings are high-rise apartments, and I bet there has been more than one numskull who's tried to light a barbecue grill on his balcony and let the fire get out of hand.  If a building has potentially flammable sandwich panels, the owners better make sure that all the fire alarms and protection systems are operational, and conducting regular fire drills might not be a bad idea either.  But owners will be reluctant to advertise the fact that their building is a giant firework waiting for someone to light the fuse.

We can also be thankful that U. S. building codes flat-out prohibit the use of sandwich panels in high-rise structures.  Yes, it forces builders to use more expensive materials, and drives the cost up compared to construction costs in other countries.  But we've had enough towering infernos in this country to last us a long time, and we don't need any more.

Sources:  I referred to a Reuters report by Andrew Torchia carried on Jan. 2, 2016 on the Yahoo News website http://news.yahoo.com/dubai-blaze-raises-questions-over-gulf-skyscraper-design-160747983--finance.html#.  The architect's chatroom with a comment about the U. S. prohibition of sandwich panels and photos of similar fires in other countries is at http://www.skyscrapercity.com/showthread.php?t=1801571.  A report of a sandwich-panel-fueled fire in a Melbourne building in 2014 appeared on the Australian website http://www.architectureanddesign.com.au/news/non-compliant-cladding-fuelled-melbourne-apartmenton Apr. 28, 2015.  I also referred to the Wikipedia article on sandwich panels.