Imagine that you are an engineer working for a company, and your job is to develop a UAV. You assemble the parts, program the software, and begin testing the UAV before handing it over to the customer.
Testing the full UAV could be done by a single UAV- operator, who will take a control unit to the airfield and remotely put the UAV into action, fly it, and give you feedback.
Until few years ago, this was the standard practice, but gradually, the maturing UAV industry realized the evident drawbacks of that initial approach: there was always some level of uncertainty whether the tests were complete enough to cover all the situations.
The industry, the developers and the customers demanded a more extensive set of tests prior to the actual tests in the field.