The following is my list of some of the key guidelines, tools and rules of Design Thinking. They're for design solutions in which collaboration with stakeholders is important; when outcomes are significant; and the behaviour of the outcomes resulting from the design are dynamic. They apply from 1H to 7H design solutions (High levels of complication, high stakeholder investment in success, high risk, high significance, high cost, high technology, high complexity in outputs and outcomes). The list has been accumulated over design projects, research and reviews of tricks, tools and tips of design thinking since the 70s. The list is incomplete and additional items will be added over time.
1. Quantify whenever possible – especially in situations involving collaboration with stakeholders.
2. Abstract until it ceases to be useful - abstract the characteristics of designs and situations to inform a design and model them. Then abstract the characteristics of those characteristics to inform the design, then abstract the characterstics of those characteristics. . . . repeat until it ceases to be useful.
3. Make everything as simple as possible and no more - more importantly, avoid making things simple for simplicity sake. Use Ashby's Law to identify the limits of simplicity that will work.
4. Ensure as much of a design as possible is designed to be orthogonal in terms of its functionality. By keeping functions independent of each other it is possible to design better outcomes (see Suh's Axioms).
5. Minimise the number of different components, and the number of components overall. Design for ease of manufacture.
6. Distinguish between ‘outputs’ and ‘outcomes’, and the aspects of a design that create them. Address them differently in the design activity and collaboration. Conflating these is asking for design failure.
7. Identify the required dynamics of the behaviours of outcomes and outputs. Outcomes on any non-trivial design are dynamic and change over time. There is usually no fixed design outcome. The idea of a fixed design outcome is a fiction of simplistic design education. Often the behaviours (plural) of design outputs are also dynamic.
8. Identify the dynamics of the causal relationships between dynamics of behaviours of outputs and outcomes. Unless the designer(s) can accurately predict the ways that features of a design cause the intended and unintended dynamic behaviours of design outcomes and outputs, the designers are simply guessing - and potentially financially responsible for those guesses and any problem outcomes (Ralph Nader onwards).
9. Identify maximum decomposition possibilities from function, structure, informatic, use, sales and manufacturing analyses. Identify reasons that limit decomposition in any area of a design. Knowing the decomposition behaviours of a design space gives a lot of information about the extents of the range of possible designs, likely areas of optimal/better design solutions, possibilities for simplification of the design, design integration, and how a design will comport with other designs and technologies.
10. Identify where any significant parts of the design do not follow associative, commutative, and distributive behaviours. If so, avoid decomposition and attempts at orthogonal design in those areas because they require them. Take care not to unthinkingly presume the opposite, that areas of a design space do not follow associative, commutative, and distributive behaviours. Doing so loses the advantages of decomposition and orthogonality unnecessarily.
11. Identify significant feedback loops shaping behaviours of outputs and outcomes. The presence of feedback loops affecting the behaviours of outputs and outcomes of a design means that they will be dynamic and for more than 2 feedback loops, impossible to predict 'in mind', by drawings or via collaboration and discussion. Any attempt to be creative, use sketches or use collaborative discussions and stakeholder discussions will flounder because of a lack of ability to predict the consequences of design decisions.
12. Where number of feedback loops involved in shaping either the behaviour of outputs or outcomes is 2 or greater use system dynamic modelling – mental models, sketches and discussions between stakeholders will fail. Instead, gather information form stakeholders, create computerized system dynamics models on the basis of that information, run the models and use them to enable stakeholders to see the consequences of design decisions.
13. When using DfX approaches in design thinking, identify whether DfX implications are decomposable and/or functionally orthogonal. Where decomposition and orthogonality admit, the different forms of DfX optimization can be applied independently. This reduces compromises, yet does not always guarantee the best solution. It is limited by the usual problems of local sub-optimisation.
14. Identify the characteristics of the bounds on the solution space. By identifying the characteristics of the bounds on a design's solution space it is possible to identify design solutions faster. Where appropriate design solutions can be found by a combination of morphological analysis and brute force search.
15. Identify characteristics of the n-dimensional shape of solution space with respect to favoured dynamics of the behaviours of outcomes. This offers the opportunity to go beyond the traditions and current obsessions with focusing primarily on design outputs with outcomes secondary to outputs. Instead it provides a path to enable preferred outcomes to be used to identify the outputs that will create them (think Jeopardy and IBM Watson).