As a scuba diver I often noticed, and always was fascinated by, strange shape and movements of boxfish. These little trunks' maneuverability is incredible, they move fast in and out of reef crevices, hover in any orientation, zoom and stop momentarily. As it turned out, there is a reason for it. I found the answer in the Natural History Magazine article. A reader can find more about it by following a DaimlerChrysler link in that article. The design and its ingenuity are described there very clearly. I just want to emphasize TRIZ principles refuted by it.
Principle 1. Un-Segmentation. The boxfish success achieved using rigid, uninterrupted, un-fragmented design.
Principle 2. Un-Taking Out. All parts kept together in the successful boxfish design.
Principle 3. Un-Local Quality. The successful structure is highly uniformed.
Principle 4. Un-Asymmetry. Symmetry is an important factor in this design.
Principle 14. Un-Spheroidality. It's a box!
Principle 15. Un-Dynamics. The structural elements are immovable.
The boxfish design quite obviously contradicts many basic assumptions of TRIZ. Not only The 40 Principles, but also the basic laws: no "dynamic control", no "macro-micro", no "mono-bi-poly", etc.
The solution also is not based on the Ideal Final Result axiom: the result is not achieved "by itself", but rather a special tool is used to achieve it, i.e. the box: a clever set of immovable strakes, shaping the body from stem to stern, lets geometry, and fluid dynamics, do all the work "for free".
Thanks to Mercedes-Benz engineers, this design cannot be disregarded as simply an example of biology not following the same laws as technology. The boxfish solution is excellent - their survival is the best proof - and TRIZ is unable to find it. Moreover, TRIZ would lead the boxfish in the wrong direction, if the boxfish used TRIZ. Fortunately for the boxfish, they did not!