A fundamental challenge in IoBTs is to offer assurances on outcomes. This task develops foundations of dependability in IoBTs that allow offering high assurances in largely uncertain environments. The need for assurances cuts across all three research areas. Asset composition, system adaptation, and intelligent services must all guarantee that desirable results be achieved. Offering such guarantees is notoriously difficult because they are predicated on assumptions regarding the state of the IoBT and its environment. This state is unpredictable. Often trade-offs arise between robustness (ability to performs well across a large set of conditions) and optimality. Systems that make more assumptions about future states perform better when these assumptions are met but are more brittle as their performance is predicated on more (assumed) condition. Foundations developed in the proposed work enable design of systems that are dependable and robust without compromising performance.