onboard intelligence

Onboard intelligence is the use of artificial intelligence (AI) that is housed locally in the device it operates.

The use of onboard intelligence helps accommodate AI uses where offloading resources externally is problematic. Such situations include remote environments and ones where connectivity or power delivery is an issue. While onboard intelligence is often web-accessible, it is often focused on autonomous operation capabilities.

Often AI implementations use both local and remote computing resources to enable greater processing power. This separation of resources can be to reduce local power draw. They may also be made separate for mobility or battery life considerations.

Mobility is a major concern in automotive applications of onboard intelligence. As vehicles can’t rely on the ability to connect adequately for the operation of remote compute power, autonomous driving cars operate with onboard intelligence. The localization of AI through onboard intelligence enables reliability in split second, life-or-death decisions that the upcoming technology has to make.

On the other hand, this divided approach limits where AI can be used. NASA, for example, has created an onboard intelligence solution for space called Frontier. Frontier is what is called an intelligent decision engine. While remaining simple enough to house locally onboard intelligence, Frontier can learn to make better decisions based on experience. This learning capability makes Frontier and other onboard intelligence more adaptable to uncertain environments such as space or other.

Commercially used drones may also benefit from onboard AI in order to enable autonomous operations. Even in crafts not operated autonomously, onboard intelligence can be a back up to help ensure the craft doesn’t fail its delivery or mission in the event of lost connectivity.