CES 2020.

OUR GOAL.

   To build an entire city made out of LED, that is fully interactive as well as a full body experience for fully autonomous drones that will live and work within this city.

 

TECH SPEC'S.

Over (1,110) LED panels was used.

Over (275) square meters / almost (3,000) square feet of LED was used.

Over (30) million pixels was rendered in real time.

Over (10) 4K LED processors was used.

Over (10) custom media servers running Resolume Arena was used.

Each media server had (2) Nvidia Quadro P500 GPUs.

Over (70) 20A circuits of power was used running at 208v

Over (900) amps of voltage at 208v was used. That is over (195,000) watts of power.

About (150) home run Cat6 cables was used to connect the LED panels to the 4K processors.

 

THE CONCEPT.

   A client reached out with a vision to showcase their concept future city. This city is what they envision current cities to be in the next 10 years, so they wanted to recreate their vision into a physical reality people could actually understand and interact with.

   The client wanted to showcase a fully autonomous drone that would act primarily as a form of public transportation that is obtainable to use for your everyday working class citizen.

   There where tablets places around this city so guests of the booth could come in and schedule a flight to and from various buildings in the city (like our current ride share phone applications work today). From there, the drones AI system would schedule and route the flight in a way that benefits multiple customers.

With current cities growing, and growing on top of infrastructure that was designed almost 50 years ago, traffic can be a real problem. They found it silly that in 2020, we as humans, still primarily only used transportation on a 2D plane. Sure, we have trains and airplanes, but the majority of transportation takes place on the road with cars and buses.

They wanted to introduce a new form of public transportation at a lower altitude than airplanes, and at a higher altitude than cars and buses. This system would be governed by a fully autonomous AI that monitors and controls all aircraft's in real time. This will prevent delays in transportation as well as maintaining the safety and security of its passengers.

   These drones can be outfitted to help out in medical emergencies as accidents as well. The drones will also be sending back real time analytics so any other drone on that network can get up to date wind speeds, weather reports, and traffic updates. This vision is something they want to see come to happen in the next 5 to 10 years. Currently there are over 5 larger cities in the US already in discussion on to how they can make this happen now.

To take this city to the next level, we even added in scent machine (about 10) throughout the city, so a coffee shop would smell like a coffee shop, and when you are buy the residential zone, you would smell cut grass. The ocean would smell like an ocean. We really wanted to make this a full body experience.

 

 

THE DESIGN.

   In order to build a city, we needed to first design the infrastructure of what it will contain.

Like all major cities, it needed residential areas, schools, airports, hospitals, grocery stores, office buildings, etc. We had to get a building count that would demonstrate what a footprint of a larger city would feel like in the space allocated for us to build in. We ended up with a city that had over 25 buildings in the 80 square meter / 860 plus square feet of LED floor.

The majority of the LED panels used where 2.5mm. With some 3.9mm thrown into the mix as well.

   We wanted each media server to have its own source local to the machine to reduce any points of failure a networked source might incur.

The entire city was a real time AI environment where the people interacted with each other and real life events took place. The drones would then use the cities activities to take into account when directing the drones flight path.

This city is designed and ran off a local source of Unreal engine on each one of the media servers.

The texture's generated in Unreal is then routed into Resolume Arena using Spout. Resolume Arena then takes the texture and maps it to the buildings. We also used Resolume Arena for color correction on the fly.

Once all of the buildings have been mapped and the north, south, east, and west orientation of each building has been set, we then took a snapshot of the city during the daytime. I then loaded that image into each of the 4K processors as a pre-stored screen image, in case the signal to the 4K processors gets disconnected, the 4K processors will automatically switch to this image.

Since the city is running a full 24-hour simulation in a time frame of 1 hour, I wanted to take short videos (about 5 min each) for morning, afternoon, evening, and night. These videos were placed next to the Spout source and was to be used as a backup system, in case the Spout source was to ever drop out due to Unreal errors or crashes.

   Inside Resolume Arena’s Advance Output, I created 25 sample points per building. I sampled the RGB values of these building (25 times per building) and outputted those RGB values via Artnet (DMX) into a custom application I wrote in TouchDesigner. This program would sum up the RGB values per building, and then with threshold limits I could define, I could use those RGB values as a simple logic switching system, that would detect if the Unreal Spout source has dropped out or not. This application lived remotely off the media servers on a laptop and was used for real time monitoring of all the buildings in the city, as well as a triggering application for certain weather and emergency events. If at any point the Unreal Spout source would drop out (go black) rather than the city blacking out, my application would then trigger an OSC command to trigger the video clips we programmed into the Resolume Arena composition. Once the Unreal Spout source would come back on, the application would automatically switch back to the live Spout source. This system ensured there would be no downtime in the city. This application also synced up to the Unreal clock, so if building 3 dropped out at night, the video file recorded at night would trigger for building 3 (specific IP’s were assigned for accurate control). With this backup system in place, as well as the 4K processors being programmed with a pre-stored screen (upon signal loss), the city should never go black unless we tell it to.

The booth’s show controller wanted to call out a few audibles during the day, such as a weather emergency or a vehicle crash that the drones could step in and help medivac the victims to the hospital. These commands were sent out in the same application as the real time monitoring back up system application, as well as programmed into a tablet (via OSC) for remote access.

   Each server was configured one for one, with the only differences being which buildings where enabled in the output map of the Unreal engine, and inside of Resolume Arena’s advanced output, we had all of the buildings/4K processors mappings inside a single file. This way, if a server failed, we could unplug the Display Port cable from that server and plug it into a different server and then just turn on the crashed server’s building textures in Unreal, and then turn on the crashed server’s advanced output screen (in addition to the current output screen) and we can have that portion of the city back up and running in under 1 min (all the while a pre-stored image would be in place so no one would know the server was down). There was a lot of thought put into place when it came to the city design. Each 4K processors output Ethernet port was not loaded any more than 92% in total bandwidth. Each circuit of power was not loaded more than 90% in total load capacity (keep in mind that the total load is only when the entire circuit has an image of full white, which the city would never reach). We had a backup for a backup system in place. We had real time monitoring of each building input source, and real time monitoring of each LED panel’s voltage and temperature. There never was a time we didn’t know exactly what was going on and what problems we could expect to see in the upcoming day.

   In the back of the booth, we had a large presentation wall for keynote speakers and that was controlled by another Resolume Arena server. There was also a confidence monitor for the speakers notes being ran off this server. With our lighting director giving the entire booth a sense of time of day with color temperature and brightness, and lightning when the storms came in, to our sound engineer providing over 30 zones of audio for each specific area of the city, the overall outcome of this build came out beautifully. To take this city a step further, we even added in over 10 scent machines for various locations inside the city. The coffee shop smelled like a coffee shop, the residential area smelled like fresh cut grass, the ocean smelled like an ocean. The client even had a one to one scale replica of the drone that would be used. It allowed for booth attendees to set inside and see what it would be like in real life.

THANK'S.

   It was an absolute blast to be a part of our client’s vision and help bring it to life. It certainly has been one of the more complex setups I have been involved with, and by far, the most organized. We had detailed folders for each building showing cable wiring diagrams, to pixel maps, so everyone could stay on the same page when working on our individual roles to build this city. Below are some pictures taken to give you a better grasp of what we accomplished.

   To all of the team members involved in this impressive production, I say bravo for a job well done. Thank you doesn’t cover the amount of appreciation I have for all of you. Great work team!

 
 
 
 
 
 
 

_