Cancer's complexity has eluded conquest by reductionist science for over a century. Globally, cancer is a resource-consuming epidemic. A consensus has emerged that a more integrated and multi-scale approach is necessary, but making cutting-edge physical science compatible with oncology research has required cognitive shifts and innovative approaches to managing and stimulating interdisciplinary work. Mechanics, spatial organization, and geometry all play an important role in cancer. Mathematical modeling has yielded new possibilities in treatment design. Advances in imaging, microscopy, and nanotechnology have great potential to be converted into better diagnostic and treatment technology. The overflow of cancer data requires the application of information science beyond "omics." Truly interdisciplinary collaboration between scientists whose training backgrounds so widely differ poses a number of intellectual and communication challenges. Huge quantities of information and very different kinds of expertise have to be dynamically managed, striking a balance between big picture visionaries, basic scientists, and efficiently improving clinical practice. Panelists will discuss their work in physical oncology collaborations, talking about results, modes of scientific thinking, challenges they have faced, and techniques to increase interdisciplinary productivity. They will touch on ways educators and policy-makers can support interdisciplinary collaboration.