I studied engineering and we used the term robust in a more casual parlance when describing something that remains fit for purpose in the face of numerous (engineering) changes and within a wide-ranging tolerance. But it was always a comparative ("this is more robust than that" etc).
A design that is ultra-sensitive to its operating parameters and only worked, say, in temperatures of 10 to 11 deg C, would not be as robust as a similar design that works in a wider range, for example.
Likewise a design that had no redundancy and no fragile parts would not be as robust as one that had redundancy and could be dropped on its head, etc.
Also, a prototype might be considered a robust design if it is more adaptable to subsequent required changes than a prototype that can not be altered in any way.
But there's no single Engineering definition as far as I know.