Launchscene
PricingSign in
Explore

Perceptron AI - Today we're releasing Mk1.5: a new intelligence layer for embodied ag

Perceptron AI · Sep 2026 · 1:53 · AI-annotated

Length
1:53
Scenes
22 · 5.2 s avg
Product on screen
0:01
Features shown
2
Call to action
1:38 · 87%
Voiceover
Yes
00:00:00 / 01:53:23

Perceptron AI’s 1:53 launch film runs 22 scenes, about 5.2 s each. The product shows up at 0:01 and 2 features are revealed. The call to action lands at 1:38 (87% in), over a voiceover. The video is built as one long split screen: live drone footage on the left and a cream Thinking panel that types reasoning in real time on the right, bookended by a live-action hook, a pixel-text product card, and a logo close. Signature moves are text resolving out of orange pixel blocks, orange highlight sweeps across reasoning lines, and the persistent two-panel layout that holds for the whole task.

Footage neededEstimated needs: 4 capture groups — your screens or recordings · 1 official logo — reused across the film · 4 filmed passages — shoot, supply or replace
Plan my film from this onePrintable storyboard

Is this your launch? to your site or press kit.

More like this

Perceptron AI - Today we're releasing Mk1.5: a new intelligence layer for embodied ag, scene by scene

· 22 scenes as text

  1. 01 · Hook · 0:00–0:01

    A white quadcopter drone rests on dirt ground in a wooded area, a folding table and a person's legs blurred in the background. The rotors spin up and the drone lifts off, drifting left and upward out of frame. The camera holds a low, static angle.

    Camera: Static

  2. 02 · Product intro · 0:01–0:22

    A man leans over an outdoor folding table with a laptop and drone while the drone hovers behind. The shot blurs and letterboxes as pixelated text types Perceptron Mk 1.5, a sans-serif subtitle, then Sees, Hears, Reasons, Acts appear in a staggered list before the image sharpens.

    On-screen text: “Perceptron Mk 1.5”, “Real-time”, “embodied reasoning model”, “Sees”, “Hears”, “Reasons”, “Acts”

    Techniques: Kinetic type

    Camera: Static

  3. 03 · Problem · 0:22–0:34

    A white drone hovers with spinning propellers beside an outdoor table where a person rests hands over a laptop and controller. A white “Voice Prompt” card slides in lower right, its caption typing in word by word, then the frame wipes to a cream panel reading “Thinking...”.

    On-screen text: “Voice Prompt”, “The gate's been stuck open since last night. Can you go see what's wrong with it?”, “Thinking...”

    Techniques: Kinetic type, Masking

    Camera: Pan

  4. 04 · Demo · 0:34–0:37

    A white drone hovers before blurred green foliage on the left. On the cream right panel, a Thinking label tops a sentence that types out word by word, a highlight box flashing over its first phrase, a small drone icon fading in, and a Voice Prompt card fixed below.

    On-screen text: “Thinking...”, “Voice Prompt”, “The gate's been stuck open since last night. Can you go see what's wrong with it?”, “Gate is not yet visible; look left along the driveway to locate it and check the approach before translating near vegetation.”

    Techniques: Split screen, Kinetic type

    Camera: Pan

  5. 05 · Demo · 0:37–0:39

    Left: live footage of a gray quadcopter hovering before green trees, drifting upward. Right: a cream panel with an animated icon beside 'Thinking...', a reasoning line with 'Gate is not yet visible' in an orange highlight, an orange wireframe drone and a descending dot, plus a Voice Prompt card.

    Techniques: Split screen

    Camera: Static

  6. 06 · Demo · 0:39–0:49

    A split layout pairs a live drone view of a driveway on the left with a cream panel. The panel's thinking text highlights phrases in orange, then swaps to a new line as a wireframe drone icon tilts and an orange arrow traces its turn.

    Techniques: Split screen

    Camera: Push in

  7. 07 · Demo · 0:49–0:51

    A split-screen layout pairs a forward-moving forest video on the left with a cream panel on the right. The panel holds a Thinking label, a paragraph with orange-highlighted phrases, a wireframe mesh with an orange arrow, and a Voice Prompt card.

    Techniques: Split screen

    Camera: Static

  8. 08 · Demo · 0:51–0:53

    A split layout pairs a forest video feed on the left with a cream panel on the right. The feed cuts from dense trees to an elevated driveway view as the panel's reasoning text updates with an orange highlight. A white voice prompt card anchors the bottom right.

    Techniques: Split screen

    Camera: Push in

  9. 09 · Demo · 0:53–0:55

    A split-screen view shows a high-angle video of a garden on the left and a text interface on the right. The text panel updates with new instructions below a 'Thinking' status, while a voice prompt box remains static at the bottom.

    On-screen text: “Thinking...”, “An elevated over-shrub approach: advance 2 m, developing toward the open lawn, keeping branches and the 6.5 m front return in contact limit.”, “Advance to the shrub edge, looking over the open lawn. Position: 4 m ahead of the contact limit near the 6.5 m front return.”, “Voice Prompt”, “The gate's been stuck open since last night. Can you go see what's wrong with it?”

    Techniques: Split screen

    Camera: Static

  10. 10 · Feature reveal · 0:55–1:09

    A split view shows a drone feed of a wooded driveway on the left, orange gate-door boxes tracking the metal gate as the camera advances, while a right panel streams navigation instructions word by word above a static voice prompt card.

    Techniques: Split screen, Kinetic type

    Camera: Push in

  11. 11 · Feature reveal · 1:09–1:13

    Split screen: left, a shot of a black metal gate on a driveway pulls back to reveal a yellow fire hydrant; right, a cream panel types out new Thinking instructions above a static voice prompt card.

    Techniques: Split screen, Kinetic type

    Camera: Static

  12. 12 · Feature reveal · 1:13–1:24

    Left half holds a still photo of a black metal gate with opener arm and hinge box; right panel types the assistant's thinking notes word by word, then flips from light to dark cards showing a focus crop, an image search with result thumbnails, and a web search.

    Techniques: Split screen, Kinetic type

    Camera: Pull out

  13. 13 · Feature reveal · 1:24–1:25

    Two phone screens sit side by side; the left dark panel's search text settles into place while the right light panel holds a static blog article about a gate opener, a cursor resting at its lower corner.

    Techniques: Split screen

    Camera: Pan

  14. 14 · Feature reveal · 1:25–1:27

    A split view shows a dark AI assistant panel on the left streaming search results and explanatory text, while a white article page on the right scrolls down through step-by-step troubleshooting headings.

    Techniques: Split screen, Kinetic type

    Camera: Pan

  15. 15 · Feature reveal · 1:27–1:28

    A dark assistant panel sits on the left showing a Thinking for 11s header, five source links, and a paragraph of diagnostic text, while a faded JAY360 support page with a gate switch photo sits behind it on the right. The panel holds still as the timer ticks.

    Camera: Static

  16. 16 · Demo · 1:28–1:29

    A dark AI-agent card on a cream background scrolls its source list and body text into place, then slides left to reveal a white diagnosis card beside it. The white card states a gate-cable fault and lists the model, with a sources pill and icons below.

    Camera: Pan

  17. 17 · Demo · 1:29–1:32

    Split screen on a cream background: left, a live-action clip of a black automated gate in a leaf-strewn yard; right, a black card with a white AI diagnostic report that fades in. The report lists model, issue, cause, fix, and confidence, then a sources pill and feedback icons.

    Techniques: Split screen, Kinetic type

    Camera: Static

  18. 18 · Demo · 1:32–1:34

    A split screen pairs a robot camera panning from a black metal gate beside a yellow hydrant up into a pine forest on the left with a dark diagnostic report panel on the right listing the gate fault and fix.

    Techniques: Split screen

    Camera: Static

  19. 19 · Demo · 1:34–1:36

    Split screen on a cream background. Left: an aerial drone shot of a white house among pines, the camera tilts down and pushes in to the roof. Right: a dark report card with a diagnosis line and a bulleted summary, held static.

    Techniques: Split screen

    Camera: Static

  20. 20 · Demo · 1:36–1:38

    Split screen: a low-angle robot camera feed of a forest path drifts forward on the left while a dark report card on the right lists a gate diagnosis, cause, fix and confidence. The whole frame fades to pale cream at the end.

    Techniques: Split screen

    Camera: Static

  21. 21 · CTA · 1:38–1:46

    A UI report card fades out to a plain cream field. Centered dark text builds phrase by phrase, each resolving out of orange pixel blocks: “Mk 1.5.”, then “Any task.”, “Any embodiment.” and “Available today.” The full line holds, then fades out.

    On-screen text: “Mk 1.5.”, “Any task.”, “Any embodiment.”, “Available today.”

    Techniques: Kinetic type

    Camera: Static

  22. 22 · Outro · 1:46–1:53

    On a cream field, a thin black wave mark draws in at center, then the word Perceptron resolves to its right, the final letters fading up from a pixelated state. The lockup holds, then the frame clears to blank.

    Camera: Static