Edge AI for ground & mobile platforms
On-platform reasoning for robots that work where the network does not.
Real-time SLAM, multi-sensor fusion, and language- or vision-model reasoning — running directly on the robot, under shock, vibration, and full mil-spec temperature swings. No tethered GPU. No cloud round-trip.
Use cases
Where on-platform AI earns its keep
Real-time SLAM & localization
Visual-inertial SLAM, LIDAR fusion, and odometry running on-platform with deterministic latency — no offloading to a chase vehicle.
Multi-sensor perception
Camera, LIDAR, radar, and event-camera streams fused on-device. Detection, segmentation, and tracking side-by-side with a language model.
On-platform reasoning
Run a vision-language model on the robot. Plan from scene context, follow free-form instructions, and ground them in what the camera sees.
Manipulation & control
Closed-loop policies, grasp planners, and control networks executed on-platform with the latency budget the actuator needs.
Comms-denied autonomy
Decisions stay on the robot when the link drops. No model degradation when the uplink does.
Mil-spec environment
Operate from −40 to +85 °C, MIL-STD-810H shock and vibration, IP67 ingress, secure boot and per-module attestation.
The matched module
TerraBot X
FPGA-based · robotics-grade
TerraBot X is built for ground robots, mobile platforms, and autonomous vehicles — running real-time SLAM, multi-sensor fusion, and language- or vision-model reasoning directly on the device, under shock, vibration, and full mil-spec temperature swings. Built on the Invotet Unified Engine running in an FPGA fabric you can buy today — not a chip-down ASIC waiting on tape-out.
- Throughput
- 38 GOPS
- Power envelope
- 4.5 W
- Operating range
- −40 to +85 °C
- Environmental
- MIL-STD-810
Why this module, this vertical
The properties that map to your platform.
Up to 20× efficiency
A unified compute engine — systolic and vector processing in one — purpose-built for transformer workloads. Smallest logic footprint, highest utilization, up to 20× more efficient than NVIDIA Jetson.
Sustainable autonomy
Frontier-class models inside a sub-15W envelope. AI fits inside the battery or solar budget — Size, Weight, and Power optimized for every module.
Transformer-grade fidelity
BF16-native execution preserves training-equivalent accuracy at 95% sustained utilization, with native flash attention and hardware tensor-parallel sync. A cycle-accurate hardware trace buffer and compile-graph-to-hardware specialization make every inference verifiable and tuned to the workload.
GPT-native logic
Matrix multiplication, softmax, element-wise operations, and the rest of the transformer operator set run natively in purpose-built logic — no general-purpose emulation tax.
Invotet SDK
Compile once. Deploy to every Invotet module.
A unified Python SDK that ingests PyTorch, ONNX, and HuggingFace checkpoints, quantizes for Invotet modules, and ships a deterministic runtime to the device. No CUDA in the loop.
Framework
PyTorch
Trace or torch.export checkpoints compile directly with no rewrite.
Framework
ONNX
Standards-based interchange — compile any ONNX-exported model.
Framework
HuggingFace
transformers checkpoints land on Invotet through a one-line loader.
Talk to the team that ships the modules.
Most robotics conversations start with a sample unit and a workload. Tell us what your robot has to think about and we will line up a TerraBot X eval kit and route the right datasheet the same day.