The DJI manifold has been presented as the new revolution for drones, a new concept for 500$ which does exactly the same as Erle-Brain 2 (217$) does, but doubling the price.
This new board is nothing new when compared to other existing embedded platforms including the sole novelty of having support for Snappy Ubuntu Core, something that many others have done already. Manifold seems to be a Jetson TK1-like board with slight some modifications and extra I/O (note that the Jetson TK1 costs $192 https://developer.nvidia.com/embedded/buy/jetson-tk1-devkit).
When compared to Erle-Brain 2 -the new artificial brain to make robots and drones from Erle Robotics-, briefly we note that:
- although Manifold has a relevant amount of computational resources, the device is not capable of being an autopilot thereby it needs to come together with another solution.
- Its power consumption is 5 times what Erle-Brain 2 requires
- Manifold doesn’t include any onboard sensor while Erle-Brain 2 has more than 10 sensors together.
- Erle-Brain 2 powers different drones with different frames and configurations and it even goes beyond flying vehicles giving life to boats, rovers, spiders, …
Manifold is only valid for the DJI Matrice ($3500+) drone and it has a non-official USB connector.
- Erle-Brain 2 can have an onboard camera and is officially supported by the Dronecode Foundation.
- Price-wise, there’s a big difference and if anyone were to use a good solution with more computational power than the one offered by Manifold would highly likely make use of a combination of a ready-to-fly solution such as Erle-Brain 2 with something like a Jetson TK1.