Product Design and Work Systems

Time study standard time

Time study is a work measurement technique in which the time taken to perform a task is observed, recorded, and converted into a standard time. The observed time is the arithmetic mean of the recorded cycle times, \bar{t}=\frac{\sum t_i}{n}. Because the observed worker may work faster or slower than a qualified normal worker, the observed mean is adjusted by the performance rating R, giving the normal time NT=\bar{t}\times R. The performance rating is therefore the ratio of normal time to observed mean time, R=NT/\bar{t}, and a rating of 100 % means the worker is working at the standard pace. Given a standard time and an allowance, the normal time is obtained by subtracting the allowance, so the rating can be back-calculated from the standard time.

Allowances cover fatigue, personal needs and unavoidable delays. They are expressed in two common ways, and the difference matters. When the allowance fraction A is taken on normal time, ST=NT(1+A). When it is taken on standard time, ST=NT/(1-A), which gives a larger standard time for the same percentage.

• Allowance on normal time — ST=NT(1+A), where A is the fraction of normal time added.
• Allowance on standard time — ST=NT/(1-A), where A is the fraction of the standard time that is allowance.
• Shift-based allowance — when the lost time is given as a share of the whole shift, the allowance fraction is the lost time divided by the shift length, and ST=NT/(1-\text{lost time}/\text{shift length}).

When several elements are timed, each element's observed mean is multiplied by its own rating, the normal times are summed, and the allowance is applied to the total. The standard output per hour is then 60/ST when ST is in minutes. Example: observed mean 50 s, rating 110 %, allowance 10 % of normal time gives NT=55 s and ST=60.5 s.

Product development process stages

Product development is described as a sequence of stages that moves from an abstract requirement to a manufacturable product. In the systematic approach of Pahl and Beitz, design proceeds in the order conceptual design, embodiment design, and detailed design. Conceptual design establishes the function structure and the principal solution, embodiment design fixes the layout, form and materials, and detailed design produces the final dimensions, tolerances and production data. The order always goes from abstract to concrete, so detailed design cannot precede the layout that it details.

Within these phases, the main activities are often named product synthesis, analysis, simplification and standardization. These terms are easily confused, so their meanings should be memorized in pairs:

• Product synthesis — generation of the conceptual solution that answers the design brief, linked to design conceptualization.
• Product analysis — conversion of the conceptual design into an engineering-science-based model so that performance can be calculated and tested.
• Product simplification — reduction of the number of parts without losing any function, for example by combining or eliminating components.
• Product standardization — maintaining uniformity and consistency of parts, materials and specifications.

Concept selection is a step inside conceptual design. Its usual steps are preparing a selection matrix in which concepts are scored against weighted criteria, ranking the concepts by their scores, and then combining and improving promising concepts before one is chosen. Developing a cost model of the product is a separate estimation activity, so it is not a step of concept selection.

Work sampling observations

Work sampling is a random-observation technique that estimates the proportion of time a worker or machine spends in each state, such as working or idle. Each observation is a snapshot at a random instant. If x of the n observations show the worker idle, the estimated idle proportion is \hat{p}=x/n, which estimates the fraction of time spent idle.

Because the number of observations in a given state follows a binomial distribution, the standard deviation of the estimated proportion is \sigma_{\hat{p}}=\sqrt{\frac{\hat{p}(1-\hat{p})}{n}}=\sqrt{\frac{x(n-x)}{n^3}}. The standard deviation shrinks with the square root of n, so more observations give a more precise estimate.

For a large number of observations, the binomial distribution is well approximated by the normal distribution, provided np and n(1-p) are sufficiently large. The normal approximation is therefore the basis of the sample-size formula. For relative accuracy s, meaning an error of \pm s\,p, the required number of observations is N=\frac{z^2(1-p)}{p\,s^2}. For absolute accuracy e, it is N=\frac{z^2p(1-p)}{e^2}. The value of z comes from the confidence level. A confidence of about 68.27 % corresponds to z=1, about 95.45 % to z=2, and about 99.73 % to z=3. A pilot study provides only the estimate of p, which is then inserted into the formula. Example: with p=0.2, s=0.05 and z=2, N=\frac{4(0.8)}{0.2\times0.0025}=6400 observations.

Concurrent engineering

Concurrent engineering, also called simultaneous engineering, is an integrated approach in which product design, process design, manufacturing, assembly, quality and support functions work in parallel through cross-functional teams, rather than passing the work from one department to the next. Its central idea is that all elements of the product life cycle, from conception to disposal, are considered during the product design stage. Issues of manufacturability, assembly, serviceability and recyclability are therefore resolved before tooling is made.

Because downstream activities start before upstream activities are complete, the development activities overlap. This parallel execution shortens the overall lead time for product development and reduces the number of late design changes. Concurrent engineering is therefore an integrated, not a non-integrated, approach, and it does not carry out functions strictly in sequence.

In contrast, sequential or 'over-the-wall' development completes each function before the next begins. This lengthens the lead time and tends to increase rework, because problems found by manufacturing are returned to design. For example, if a toolmaker joins the design team at the concept stage, a feature that is expensive to machine can be changed before the drawings are released.

Therblig symbols

Therbligs are the basic elemental motions of manual work, identified by Gilbreth and used in micromotion study. Each therblig has a standard abbreviation and a fixed graphical symbol, and an analyst records a cycle by writing the symbol for each motion. The symbols must be memorized as a set, because similar-looking symbols represent different motions. For example, the rest-to-overcome-fatigue therblig (RO) is a distinct symbol and must not be confused with position (P) or preposition (PP).