From Optimizing to Learning: A Development of Systems Thinking for the 1990s

In any subject concerned with rational intervention in human affairs, theory must lead to practice; but practice is the source of theory: neither theory nor practice is prime. We can examine this ‘groundless’ relation by asking what intellectual framework F is applied in what methodology M to what area of application A? If we do this for O.R., systems analysis, systems engineering etc., we see that F and M have changed dramatically between the 1950s and the 1980s, yielding the ‘hard’ and ‘soft’ traditions of systems thinking. The ‘hard’ tradition, based on goal seeking, is examined in the work of Simon and contrasted with the ‘soft’ tradition, based on learning, as exemplified in the work of Vickers and the development of soft systems methodology. The two are complementary, but the relation between them is that the ‘hard’ is a special case of ‘soft’ systems thinking. This analysis makes sense of the recent history of management science and helps to prepare us for the 1990s.     

The development of management sciences/operational research discourses: surveying the trends in the US and the UK

This paper explores the trends in American and British management science/operational research (MS/OR) during the last 25 years. We argue that British MS/OR has developed a soft and systemic approach to MS/OR practice, which has resulted in the emergence of a number of interpretive and critical-oriented methodologies. American MS/OR practice has remained closed to the positivistic discourse. Using a set of keywords and authors’ names associated with the main features of the interpretive and critical MS discourses, we surveyed articles published in three major US MS/OR journals. We compare these results with trends in the UK MS/OR scene. Findings appear to confirm the different directions taken by the MS/OR practice across the Atlantic. The paper posits possible reasons underpinning these differences: firstly, the particular methodological path followed by the British MS/OR, from early ‘soft systems’ applications in the early 1970s to the now well-established ‘Problem Structuring Methods’; and secondly, continuous engagement between the systems and MS/OR British communities (a dialogue that seems not to have occurred in the US). The paper contributes to a reflection on the MS/OR historical developments and contrasts these developments in both countries, two areas of OR significantly under-researched.     

Multimethodology in series and parallel: strategic planning using hard and soft OR

This paper examines two distinct ways in which hard and soft operational research (OR) methodologies can be combined, in series and in parallel. Multimethodology in series is acknowledged as the simpler and more common approach. Multimethodology in parallel is identified as having the potential to provide significant benefits to projects in political, changing, or ‘wicked’ contexts that multimethodology in series cannot. Observations regarding these approaches to multimethodology are examined in light of an information systems strategic planning project in the Australian public sector. Two distinct methodologies were combined in the project: soft systems methodology and project management. These methodologies are based on the soft and hard paradigms, respectively. However, findings in this paper have the potential to be transferred to combinations of other hard and soft OR methodologies.     

On the non-amenability of the reflective quotient

Let O(f, ?) be the integral orthogonal group of an integral quadratic form f of signature (n, 1). Let R(f, ?) be the subgroup of O(f, ?) generated by all hyperbolic reflections. Vinberg [Vi3] proved that if n ≥ 30 then the reflective quotient O(f, ?)/R(f, ?) is infinite. In this note we generalize Vinberg’s theorem and prove that if n ≥ 92 then O(f, ?)/R(f, ?) contains a non-abelian free group (and thus it is not amenable).     

Quaternion rings and octonion rings

In this paper, for rings R, we introduce complex rings ?(R), quaternion rings ?(R), and octonion rings O(R), which are extension rings of R; R ? ?(R) ? ?(R) ? O(R). Our main purpose of this paper is to show that if R is a Frobenius algebra, then these extension rings are Frobenius algebras and if R is a quasi-Frobenius ring, then ?(R) and ?(R) are quasi-Frobenius rings and, when Char(R) = 2, O(R) is also a quasi-Frobenius ring.     

Geometry of Coadjoint Orbits and Multiplicity-one Branching Laws for Symmetric Pairs

Consider the restriction of an irreducible unitary representation π of a Lie group G to its subgroup H. Kirillov’s revolutionary idea on the orbit method suggests that the multiplicity of an irreducible H-module ν occurring in the restriction π|_H could be read from the coadjoint action of H on \(\mathcal {O}^{G} \cap \text {pr}^{-1}({\mathcal {O}}^{H})\), provided π and ν are ‘geometric quantizations’ of a G-coadjoint orbit \(\mathcal {O}^{G}\) and an H-coadjoint orbit \(\mathcal {O}^{H}\), respectively, where \(\text {pr} \colon \sqrt {-1}\mathfrak {g}^{\ast } \to \sqrt {-1}\mathfrak {h}^{\ast }\) is the projection dual to the inclusion \(\mathfrak {h} \subset \mathfrak {g}\) of Lie algebras. Such results were previously established by Kirillov, Corwin and Greenleaf for nilpotent Lie groups. In this article, we highlight specific elliptic orbits \(\mathcal {O}^{G}\) of a semisimple Lie group G corresponding to highest weight modules of scalar type. We prove that the Corwin–Greenleaf number \(\sharp (\mathcal {O}^{G} \cap \text {pr}^{-1}({\mathcal {O}}^{H}))/H\) is either zero or one for any H-coadjoint orbit \(\mathcal {O}^{H}\), whenever (G,H) is a symmetric pair of holomorphic type. Furthermore, we determine the coadjoint orbits \(\mathcal {O}^{H}\) with nonzero Corwin–Greenleaf number. Our results coincide with the prediction of the orbit philosophy, and can be seen as ‘classical limits’ of the multiplicity-free branching laws of holomorphic discrete series representations (Kobayashi [Progr. Math. 2007]).     

A method for analysing operational complexity in supply chains

This paper proposes a method for analysing the operational complexity in supply chains by using an entropic measure based on information theory. The proposed approach estimates the operational complexity at each stage of the supply chain and analyses the changes between stages. In this paper a stage is identified by the exchange of data and/or material. Through analysis the method identifies the stages where the operational complexity is both generated and propagated (exported, imported, generated or absorbed). Central to the method is the identification of a reference point within the supply chain. This is where the operational complexity is at a local minimum along the data transfer stages. Such a point can be thought of as a ‘sink’ for turbulence generated in the supply chain. Where it exists, it has the merit of stabilising the supply chain by attenuating uncertainty. However, the location of the reference point is also a matter of choice. If the preferred location is other than the current one, this is a trigger for management action. The analysis can help decide appropriate remedial action. More generally, the approach can assist logistics management by highlighting problem areas. An industrial application is presented to demonstrate the applicability of the method.     

Operational complexity and supplier–customer integration: case study insights and complexity rebound

The main contribution of this paper is the demonstration that, contrary to conventional thinking, a measurable increase in the operational complexity of the production scheduling function between two companies can occur following closer supply chain integration. The paper presents the practical application of previous work carried out and validated by the authors in terms of (a) methodology for measuring operational complexity, (b) predicted implications of Supplier–Customer integration and (c) derivation of an operational complexity measure applied to before and after Supplier–Customer integration. This application is illustrated via a longitudinal case study. The analysis is based on information theory, whereby operational complexity of a Supplier–Customer system is defined as the amount of information required to describe the state of this system. The results show that operational complexity can increase when companies decide to integrate more closely, which is a fact likely to be overlooked when making decisions to pursue closer Supply-Chain integration. In this study, operational complexity increases due to reduced buffering arising from reduction in the Supplier's inventory capacity. The Customer did not change their operational practices to improve their schedule adherence post-integration, and, consequently, suffered an increase in complexity due to complexity rebound. Both the Supplier's and Customer's decision-making processes after the case study reported in this paper were enhanced by being able to quantify the complex areas to prioritise and direct managerial efforts towards them, through the use of the operational complexity measure. Future work could extend this study (in the ‘low product customisation’ and ‘low product value impact’ quadrant) to investigate Supplier–Customer integration in other quadrants resulting from further combinations between ‘product customisation’ and ‘product value impact’ levels.     

A sequential method for preventive maintenance and replacement of a repairable single-unit system

This paper is concerned with when to implement preventive maintenance (PM) and replacement for a repairable ‘single-unit’ system in use. Under the main assumption that a ‘single-unit’ system gradually deteriorates with time, a sequential method is proposed to determine an optimal PM and replacement strategy for the system based on minimising expected loss rate. According to this method, PM epochs are determined one after the other, and consequently we can make use of all previous information on the operation process of the system. Also the replacement epoch depends on the effective age of the system. A numerical example shows that the sequential method can be used to solve the PM and replacement problem of a ‘single-unit’ system efficiently. Some properties of the loss functions W(L? _n ,b? _n ) and W? _r (N) with respect to PM and replacement respectively are discussed in the appendix.     

Bloom’s inequality: commutators in a two-weight setting

In 1985, Bloom characterized the boundedness of the commutator [b, H] as a map between a pair of weighted L^p spaces, where both weights are in A_p. The characterization is in terms of a novel BMO condition. We give a ‘modern’ proof of this result, in the case of p = 2. In a subsequent paper, this argument will be used to generalize Bloom’s result to all Calderón–Zygmund operators and dimensions.     

Meromorphic Jost functions and asymptotic expansions for Jacobi parameters

We show that the parameters a _n , b _n of a Jacobi matrix have a complete asymptotic expansion

$a_n^2 - 1 = \sum\limits_{k = 1}^{K(R)} {p_k (n)\mu _k^{ - 2n} + O(R^{ - 2n} ),} b_n = \sum\limits_{k = 1}^{K(R)} {p_k (n)\mu _k^{ - 2n + 1} + O(R^{ - 2n} )} $

, where 1 < |µ_j| < R for j ? K(R) and all R, if and only if the Jost function, u, written in terms of z (where E = z + z ^?1) is an entire meromorphic function. We relate the poles of u to the µ_j’s.

     

Tauber-Konstanten für Verschiedene Tauberbedingungen bei den Kreisverfahren der Limitierungstheorie

Till now, we know Tauberian constants for the ‘Kreisverfahren’ with the conditions lim sup |n ^1/2 a _n|<∞ and lim sup |n ¹ a _n|<∞. Now, we obtain constants for the more general condition lim sup |n ^pa_n|<∞ with anyp(=∞<p<+∞). These constants are not always 0 or ∞, even if 1/2<p<1; therefore the Tauberian condition lim sup |n ^pa_n|<∞ is ‘appropriate’ for 1/2≦p≦1.     

Polyboxes,Cube Tilings and Rigidity

A non-empty subset A of X=X ₁×???×X _d is a (proper) box if A=A ₁×???×A _d and A _i ?X _i for each i. Suppose that for each pair of boxes A, B and each i, one can only know which of the three states takes place: A _i =B _i , A _i =X _i ?B _i , A _i ?{B _i ,X _i ?B _i }. Let F and G be two systems of disjoint boxes. Can one decide whether ∪F=∪G? In general, the answer is ‘no’, but as is shown in the paper, it is ‘yes’ if both systems consist of pairwise dichotomous boxes. (Boxes A, B are dichotomous if there is i such that A _i =X _i ?B _i .) Several criteria that enable to compare such systems are collected. The paper includes also rigidity results, which say what assumptions have to be imposed on F to ensure that ∪F=∪G implies F=G. As an application, the rigidity conjecture for 2-extremal cube tilings of Lagarias and Shor is verified.     

Reflections on queue modelling from the last 50 years

Queueing theory continues to be one of the most researched areas of operational research, and has generated numerous review papers over the years. The phrase ‘queue modelling’ is used in the title to indicate a more practical emphasis. This paper uses work taken predominantly from the last 50 years of pages of the Operational Research Quarterly and the Journal of the Operational Research Society to offer a commentary on attempts of operational researchers to tackle real queueing problems, and on research foci past and future. A new discipline of ‘queue modelling’ is proposed, drawing upon the combined strengths of analytic and simulation approaches with the responsibility to derive meaningful insights for managers.     

Double-bosonization and Majid’s conjecture (IV): Type-crossings from A to BCD

Both in Majid's double-bosonization theory and in Rosso's quantum shuffle theory, the rankinductive and type-crossing construction for U_q(g)'s is still a remaining open question. In this paper, working in Majid's framework, based on the generalized double-bosonization theorem we proved before, we further describe explicitly the type-crossing construction of U_q(g)'s for(BCD)_n series directly from type An-1via adding a pair of dual braided groups determined by a pair of(R, R′)-matrices of type A derived from the respective suitably chosen representations. Combining with our results of the first three papers of this series, this solves Majid's conjecture, i.e., any quantum group U_q(g) associated to a simple Lie algebra g can be grown out of U_q(sl_2)recursively by a series of suitably chosen double-bosonization procedures.     

A family of non-isomorphism results

We calculate the local groups of germs associated with the higher dimensional R. Thompson groups nV. For a given \({n\in N\cup\left\{\omega\right\}}\) , these groups of germs are free abelian groups of rank r, for r ≤ n (there are some groups of germs associated with nV with rank precisely k for each index 1 ≤ k ≤ n). By Rubin’s theorem, any conjectured isomorphism between higher dimensional R. Thompson groups induces an isomorphism between associated groups of germs. Thus, if m ≠ n the groups mV and nV cannot be isomorphic. This answers a question of Brin.     

A partition-based relaxation for Steiner trees

The Steiner tree problem is a classical NP-hard optimization problem with a wide range of practical applications. In an instance of this problem, we are given an undirected graph G = (V, E), a set of terminals \({R\subseteq V}\) , and non-negative costs c _e for all edges \({e \in E}\) . Any tree that contains all terminals is called a Steiner tree; the goal is to find a minimum-cost Steiner tree. The vertices \({V \backslash R}\) are called Steiner vertices. The best approximation algorithm known for the Steiner tree problem is a greedy algorithm due to Robins and Zelikovsky (SIAM J Discrete Math 19(1):122–134, 2005); it achieves a performance guarantee of \({1+\frac{\ln 3}{2}\approx 1.55}\) . The best known linear programming (LP)-based algorithm, on the other hand, is due to Goemans and Bertsimas (Math Program 60:145–166, 1993) and achieves an approximation ratio of 2?2/|R|. In this paper we establish a link between greedy and LP-based approaches by showing that Robins and Zelikovsky’s algorithm can be viewed as an iterated primal-dual algorithm with respect to a novel LP relaxation. The LP used in the first iteration is stronger than the well-known bidirected cut relaxation. An instance is b-quasi-bipartite if each connected component of \({G \backslash R}\) has at most b vertices. We show that Robins’ and Zelikovsky’s algorithm has an approximation ratio better than \({1+\frac{\ln 3}{2}}\) for such instances, and we prove that the integrality gap of our LP is between \({\frac{8}{7}}\) and \({\frac{2b+1}{b+1}}\) .     

Variance Reduction Techniques

Estimating real-world parameter values by means of Monte-Carlo/stochastic simulation is usually accomplished by carrying out a number ‘n’ of computer runs, each using random numbers taken from a pseudo-random number generator. In order to improve the accuracy of the estimate (reduce the estimate's variance), the most common recourse is to increase n, as the estimate's variance is inversely proportional to n. Variance reduction techniques provide an alternative to increasing n. They use statistical approaches which obtain more information from the computer runs conducted, or control and direct the pseudo-random streams to optimize the information likely to be produced by a run.     

Homogenization of elliptic boundary value problems in Lipschitz domains

In this paper we study the L ^p boundary value problems for \({\mathcal{L}(u)=0}\) in \({\mathbb{R}^{d+1}_+}\) , where \({\mathcal{L}=-{\rm div} (A\nabla )}\) is a second order elliptic operator with real and symmetric coefficients. Assume that A is periodic in x _d+1 and satisfies some minimal smoothness condition in the x _d+1 variable, we show that the L ^p Neumann and regularity problems are uniquely solvable for 1 < p < 2 + δ. We also present a new proof of Dahlberg’s theorem on the L ^p Dirichlet problem for 2 ? δ < p < ∞ (Dahlberg’s original unpublished proof is given in the Appendix). As the periodic and smoothness conditions are imposed only on the x _d+1 variable, these results extend directly from \({\mathbb{R}^{d+1}_+}\) to regions above Lipschitz graphs. Consequently, by localization techniques, we obtain uniform L ^p estimates for the Dirichlet, Neumann and regularity problems on bounded Lipschitz domains for a family of second order elliptic operators arising in the theory of homogenization. The results on the Neumann and regularity problems are new even for smooth domains.