Cognata · Autonomous driving simulation · Sensor showcase video
AI Marketing Video for Cognata: 5 Sensors Tested in One Digital Twin
Cognata needed one AI marketing video to show 5 sensor makers being tested inside its Automated Driving Perception Hub, a program run with Microsoft and AMD. Advids gave every partner the same card and a layout built for its type of sensor, then labeled 15 simulated scenes, from a dropped tire on a highway to a deer crossing at night, so a viewer can see what each sensor picks up.
Last updated
Cognata · Automated Driving Perception Hub
- Client
- Cognata
- Product
- Automated Driving Perception Hub (ADPH), a sensor simulation program with Microsoft and AMD
- Made for
- ADAS and autonomous vehicle teams evaluating sensors, and the sensor makers in the program
- Partners shown
- Lidwave, Innoviz, Bright Way Vision, Foresight and ADASKY
- Format
- Split-screen sensor simulation showcase video with on-screen labels
- Length
- 3:33
- Production timeline
- 3 to 4 weeks
- Budget
- $1,200 to $2,000
The problem: 5 sensors, 1 program, and edge cases no one stages on a real road
The Automated Driving Perception Hub lets teams evaluate ADAS and autonomous vehicle sensors in a digital twin, running on Azure and AMD hardware. The program had 5 sensor makers, and each one sees the road differently: Lidwave's depth and Doppler sensors, Innoviz LiDAR, Bright Way Vision's CMOS camera system, Foresight stereo vision and ADASKY's thermal camera. A single product demo could not show all 5 fairly.
The buyer's question is simple: what does this sensor catch when the road gets hard? A tire dropped on a highway, a deer at night, people crossing in fog. Those cases are hard to stage safely on a public road, so the video had to show each one simulated, named on screen, and seen through the sensor.
How the video walks 5 sensors through 15 edge cases, scene by scene
Every test scene carries 2 logos (Cognata top left, the sensor maker top right) and a label naming the test. Select a time to play that scene above.
| Time | On screen | What it does |
|---|---|---|
| A network graphic of linked points on dark blue. | Sets a data and simulation tone in 2 seconds before any logo. | |
| Partner card with Microsoft, AMD, Cognata and "LIDWAVE" over a wireframe car, dissolving into the first test, then a red point cloud with a camera inset: "Urban Environment, Intersection" and "Static Object Detection: Dropped Tire". | Shows Lidwave's depth output on a busy junction, then on one small object in a highway lane. | |
| Innoviz card with "INNOVIZ TECHNOLOGIES", then the same "Static Object Detection: Dropped Tire", then "Object Detection: Hanging Overload" and "Safety Scenarios: Near-Collision, Late Cut-Out". | Runs the dropped tire again through a second sensor, then adds a hazard above the road and a car cutting out late. | |
| Bright Way Vision card, then split screens of the simulated scene beside the sensor view: "Harsh Weather, Night, Crossing Deer", "Harsh Weather, Fog, VRUs Crossing", "Lighting Challenge, Night, Micromobility Users". | Puts the camera system in dark, fog and a night street with riders on small vehicles. | |
| Foresight card, then 2 camera views over depth maps and a green planned path on a point cloud: "Path Planning, Off-Road Digital Twin Environment", then "Path Planning, Quarry Digital Twin Environment, VRUs" and "Path Planning, Agriculture Digital Twin Environment, VRUs". | Moves the test off the highway into dirt, rock and orchard, where stereo depth plans the route. | |
| ADASKY card, then the scene beside a thermal view: "Lighting Challenge, Long Distance, Fog", "Lighting Challenge, Long Distance, Glare", "Harsh weather, Long Distance, Night". | Shows people and animals at range in the thermal view while fog, sun glare or dark make the normal view hard to read. | |
| "NCAP Regulation, Segmentation, Night": a color-coded segmentation view beside the thermal view, then the Cognata logo. | Ends on a test named for a safety regulation and closes on the program owner. |
Storyboard: the frames that carry the Cognata AI marketing video
Key frames from the finished video, in order. Select a time to play that moment above.
LiDAR output fills the frame, with the camera view as a small inset. 
The first sensor meets one small object in a highway lane. 
The same card returns with only the partner logo changed. 
The dropped tire is run again through a second sensor. 
A hazard above the road, not just on it. 
A safety case built around another car's late move. 
Split screen: the dark normal view beside the sensor view that shows the deer. 
Stereo depth plans a route where there is no road. 
Sun glare washes out the normal view while the thermal view keeps 2 people clear. 
The last test is named for a safety regulation.
4 decisions behind the Cognata AI marketing video
Decision 1
Run the same tire past 2 sensors
"Static Object Detection: Dropped Tire" plays for Lidwave at 0:18 and again for Innoviz at 0:33. The test stays fixed, so the only thing that changes is the sensor.
Copy it: If buyers compare options on your platform, put one scenario through each option and let them see the difference.
Decision 2
One card, 5 partners
The partner card at 0:03, 0:31, 1:18, 2:00 and 2:48 keeps Microsoft, AMD, Cognata and the wireframe car in place. Only the fourth logo changes.
Copy it: For a program with several partners, lock one template and swap a single element so viewers know a new section has started.
Decision 3
Name the edge case on screen
Every test scene has a lower-third label in the order test, condition, object, such as "Harsh Weather, Night, Crossing Deer" at 1:21.
Copy it: Write each label as test, condition and object, so a viewer with the sound off still knows what is being proved.
Decision 4
Fit the layout to the sensor
LiDAR scenes use a point cloud with a small camera inset (0:08), the camera and thermal scenes sit side by side (1:21, 2:51), and stereo gets 2 camera views, 2 depth maps and a path (2:03).
Copy it: Show your product's real output in its own form, next to what a person would see in the same moment.
From sensor outputs to a 5-partner program video
- Program map. The 5 partners were set in order, each with its own sensor type and 2 to 4 simulated scenarios.
- Wireframe car. A 3D car model was refined, textured and lit as the wireframe that sits behind every partner card.
- Partner card. One animated card was built with Microsoft, AMD and Cognata fixed and a slot for each sensor maker's logo.
- Sensor layouts. Each sensor type got its own composite: point cloud with camera inset, side-by-side views, or stereo views with depth maps and a path.
- Scenario labels. Lower-third labels and corner logos were added so each scene names its test and its sensor maker.
- Sound and final edit. Sound design and the final edit paced 5 sections over 3:33.
Scope of work:
Production took 3 to 4 weeks on a budget of $1,200 to $2,000. See the 3D video production service for how a project like this is scoped.
Questions about the Cognata AI marketing video
Why does the Cognata video show the dropped tire twice?
"Static Object Detection: Dropped Tire" runs through Lidwave's sensor at 0:18 and through Innoviz LiDAR at 0:33. Keeping the test the same lets a viewer compare 2 sensors on one hazard.
Why do Microsoft and AMD appear on every partner card?
The program runs on Azure and is optimized for AMD processors and GPUs, per Cognata's description. Their logos stay fixed on all 5 cards (0:03, 0:31, 1:18, 2:00, 2:48) while the sensor maker's logo changes.
How does the video show what each sensor sees?
Each sensor type gets its own layout: a red point cloud with a camera inset for Lidwave, a side-by-side sensor view for Bright Way Vision and ADASKY, and 2 camera views with depth maps and a green planned path for Foresight from 2:03.
Which edge cases does the Cognata video cover?
15 labeled scenes, including "Object Detection: Hanging Overload" (0:48), "Harsh Weather, Night, Crossing Deer" (1:21), "Harsh Weather, Fog, VRUs Crossing" (1:30) and "Lighting Challenge, Long Distance, Glare" (2:57).
Why does the video end on an NCAP scene?
"NCAP Regulation, Segmentation, Night" at 3:18 shows a color-coded segmentation view beside ADASKY's thermal view. The last test is named for a safety regulation, then the Cognata logo appears at 3:30.
How long did the Cognata video take and what did it cost?
Production took 3 to 4 weeks on a budget of $1,200 to $2,000, covering 3D model work, animation, compositing, effects, sound design and the final edit.
What Advids clients say
"We had a great experience with ADVIDS. We required a technical 3D video to be produced. The project ran smoothly with excellent communication."Emily Gray, Quality Manager at Pacific Hoseflex
Need ADAS buyers to see what your sensor or AI model catches?
Talk to Advids