We develop a rigidity theory for bar-joint frameworks in Euclidean d-space in which specified classes of edges are allowed to change length in a coordinated fashion that requires differences of lengths to be preserved within each class. Rigidity for these coordinated frameworks is a generic property, and we characterize the rigid graphs in terms of redundant rigidity in the standard d-dimensional rigidity matroid. We also interpret our main results in terms of matroid unions. Definition 1.3. Let d ∈ be a dimension and G a graph. Then G is generically rigid in dimension d if every generic framework (G, p) is rigid; otherwise every generic (G, p) is flexible and G is generically flexible in dimension d. If G is generically rigid in dimension d, but no proper spanning subgraph of G is generically rigid, then G is isostatic in dimension d. Rigidity matroids It is implicit in [1] and first explicitly observed and used by Lovász and Yemini in [13] that generic rigidity has a matroidal structure. Definition 1.4. Fix a dimension d and let n ≥ d. Let E n be the edges of the complete graph K n . The matroid on the ground set E n of rank d n − d+1 2that has as its bases the edge sets of the isostatic graphs with n vertices in dimension d is called the d-dimensional rigidity matroid of K n and is denoted by M d,n .The restriction of M d,n to the edges of an n vertex graph G is the rigidity matroid of G, M d (G).Lovász and Yemini define M d,n in terms of a linearization of rigidity called infinitesimal rigidity that we discuss in more detail in Section 2. This approach is now standard in the field; see, e.g., Whiteley's survey [24] for an overview on the interplay between matroid theory and rigidity problems.It is easy to check, using a randomized algorithm based on Gaussian elimination, whether a specific graph G is generically rigid in dimension d for any d and number of vertices n (a detailed analysis is in [8], but this is a folklore fact). On the other hand, except for dimensions d = 1, which is folklore, and d = 2, which is due to Pollaczek-Geiringer [18] (and later rediscovered by Laman [11]), a combinatorial characterization of the matroids M d,n is a notable open problem [20, Sec. 61.1.2, "Open problems"].
Coordinated rigidity: motivation and resultsIn recent work, Nixon, Schulze, Tanigawa and Whiteley [16] defined a generalization of frameworks that enlarges the class of allowed motions. The vertices of (G, p) are partitioned into k + 1 different classes, V 0 , V 1 , . . . , V k . The set of allowed configurations p is constrained so that for j ≥ 1 all vertices i ∈ V j have the same distance to the origin (but this distance may change), and all the vertices i in V 0 lie on the unit sphere. The allowed motions are the continuous deformations in the space of allowed configurations. The motivation for studying these types of frameworks is to interpolate between rigidity in dimension d and dimension d + 1.A model for this expanding spheres setup, which is present in [16], is based on Whiteley's coning construction [23]....