Future Tense is a partnership of Slate, New America, and Arizona State University that examines emerging technologies, public policy, and society. This article is part of Update or Die, a series from Future Tense about how businesses and other organizations keep up with technological change--and the cost of falling behind. Few passengers realize that in the airline industry, we exclusively train our pilots using simulators. When a new-hire pilot flies the real airplane for the first time, it's with paying customers in the back. To create one of our simulators, we hacksaw off the pointy end of a real airplane, put it on a 6-degree-of-freedom hexapod motion platform, and outfit it with video displays so pilots have something to look at out the front window.
I'm not talking about the toy drones that float around your house, bouncing harmlessly off walls, people, and furniture, or even then $99 ones you keep losing over the ocean. Companies like DJI have built significant intelligence into their drones and vastly simplified the apps, but flying them is still a skill. SEE ALSO: DJI's Spark drone is so small and smart, it could be a game-changer DJI gets it and, now, with partner Epson, they're trying to take the confusion and risk out of learning how to fly an expensive drone. You just must be willing to pay $700 to use it. The somewhat expensive idea is smart.
Debrief tool used in the experiment displays a video replay of the operator console (similar to this map display), and a timeline of events suggested by AEMASE for discussion during debrief. The tool also includes visualizations of entity movement over time. Navy pilots and other flight specialists soon will have a new "smart machine" installed in training simulators that learns from expert instructors to more efficiently train their students. Sandia National Laboratories' Automated Expert Modeling & Student Evaluation (AEMASE, pronounced "amaze") is being provided to the Navy as a component of flight simulators. Components are now being used to train Navy personnel to fly H-60 helicopters and a complete system will soon be delivered for training on the E-2C Hawkeye aircraft, said Robert G. Abbott, a Sandia computer scientist and AEMASE's inventor.
The approach is being applied to create automated pilots for a beyond visual range flight simulator. In this domain it is necessary that a human controller can control an agent at all levels of abstraction. We believe a multi-agent action selection mechanism provides a natural way of achieving this goal. Introduction We describe an ongoing project to build intelligent agents that can be directed in real-time. The agents are designed for a real-time simulation environment where a human controller must be able to take control of particular aspects of an agents behavior while the agent continues to act. Two challenging aspects of the problem are that the agent will be expected to continue on with other tasks as well as fulfilling the operators request and that the human operator should be able to make requests at all levels of abstraction. The agents implement simulated pilots for a beyond visual range simulator called TACSI, developed by Saab Missions and Systems. The TACSI simulator is used to test and evaluate systems and to train pilots.