Human in the Loop (or) Computer in the Loop: Augmented Intelligence and Performance

A computer-centric robotic control approach applied at a professional level will play to the inherent strengths of both humans and robots. Toyota is working on Artificial Intelligence with a novel perspective on a familiar control theory concept referred to as "human in the loop." 

Although many research organizations apply this configuration, this particular case is noteworthy because of Toyota's unique approach. Rather than a typical human-in-the-loop system or full machine autonomy, the human is only making the larger high-level decisions (infrequent and easy for humans), and the computer is making the smaller high-speed decisions (frequent and complex for humans). Will Oremus of Slate.com refers to this as “computer in the loop.” 

In many instances, the frequent low-level decisions are mathematically and physically complex, monopolize cerebral function, and can have greater consequences for the system.

Currently, humans are
 better at both understanding the sequence of operations and the underlying rationale. Conversely, computers are more adept at controlling robotic systems. Some of the shortcomings of human-controlled robotics are a function of the Human–Machine Interface (HMI), but FMC Schilling Robotics and others are at the forefront of innovations in this area. 




FMC Schilling Robotics ROV Control Console


Determining what level of abstraction end users and programmers should be operating under is becoming increasingly important for systems engineers and software developers. For roboticists developing adaptive robotic and human-supervisory control applications, Actin is the most advanced programming and augmentation tool available.



For more information, visit our Actin™, Dynamic Drive™, and ROV technology pages or follow the links above.

Control Interfaces — DYNAMIC DRIVE™