VPA for robot arms
Tested in simulationSuccess rate on 10 household tasks20 trials per task, first results
SLING AI builds VPA (Vision-Predictor-Action), control software that lets a machine look through its camera, predict the next moment and move, with no large language model in the loop. The same design steers a robot arm and a drone, and at about 70 MB it runs on the machine's own low-power board instead of a costly GPU.
Where we are: Working software in simulation, released openly as models, code and data. Tests on a real robot arm and a real drone begin in the first half of 2027.
One VPA brain
About 70 MB. Runs inside the device, no costly GPU.
What we sell, to whom, and how far along we are.
Most new robot models first turn what the camera sees into language, then reason with a large language model (LLM) before they move.
LLM-based robot AI carries 4 GB or more of weights and needs an expensive GPU.
It computes its answer one word at a time, so it responds late when an obstacle suddenly appears.
Robot arms and drones each get their own AI, and a drone can rarely carry a large model at all.
Slow, heavy and fragmented: this is what keeps robots and drones out of everyday field work.
VPA drops the language step. Four small networks run in a row, and the last one writes the whole next stretch of motion in a single pass.
No language step: 5 network calls plan the whole continuous motion.
70 MB refers to Aero-VPA's 18.13 million parameters stored at full precision. VPA for robot arms is the same size class, under 20 million parameters.
The same architecture becomes VPA when trained on robot-arm data and Aero-VPA when trained on drone data.
Success rate on 10 household tasks20 trials per task, first results
Flights that reach the goal in a simulated forest180 flights; crashes fell from 72% to 22%
Open release Model Flight records Code Paper
One VPA: the same architecture, about 70 MB, 5 network calls per decision, low-power boards.
Results so far come from simulation. Tests on a real robot arm and a real drone are scheduled for the first half of 2027.
Robot AI today either runs a large language model or, for drones, builds a map before it can plan. VPA does neither.
| Approach | Model size (parameters) | Hardware needed | How it produces motion | Used for |
|---|---|---|---|---|
| Large robot AIGoogle and other big-tech labs; LLM-based | 3.3 billion to 55 billion | High-end GPU | Turns what it sees into words, then reasons | Robot arms |
| Small robot AIOpen small models; LLM-based | About 450 million | Standard GPU or PC | Turns what it sees into words, then reasons | Robot arms |
| Map-based autonomous flightTraditional approach: builds a map first | Not applicable | Several sensors | Builds a map, then computes a route | Drones |
| SLING AIVPA and Aero-VPA | About 18 million | Low-power edge board | 5 network calls, straight to motion | Robot arms and drones, same architecture |
1/100 the size of large robot AI, or smaller.
Every decision takes exactly 5 network calls.
The same architecture controls robot arms and drones.
The safety shield cut drone crashes from 72% to 22%.
Model size is the number of parameters reported in public papers: Google RT-2 up to 55 billion, OpenVLA 7 billion, π0 3.3 billion, SmolVLA 450 million.
Two revenue streams, both from the same VPA software.
Demand for robots and drones is rising fast, and the cost of high-end GPUs is what holds deployment back.
One VPA lets us go after both markets at the same time.
Sources: Korea Ministry of Land, Infrastructure and Transport, 2024 Drone Industry Survey; IMARC.
2026Done
First half of 2027
Second half of 2027
2028
AI control expertise grounded in mathematical proof.
Founder and CEO
Ph.D. in Informatics, Nagoya University (2026)
“Innovating AI control by combining theoretical rigor with hands-on field experience.”