22-Mec-B5 Product Design and Development · December 2016

Question 3 of 7: Radical and Incremental Design

Nivaar worked solution (AI-drafted; not reviewed by a licensed engineer)

Question 3: Radical and Incremental Design (15 marks)

Question text not reproduced: the examination questions are © Engineers and Geoscientists BC. Open the official past paper (linked at the top of this page) to read the question, then follow the worked solution below.

A — Two examples of each (8 marks)

The distinction is not about how much effort the change took or how new it looks. A change is radical when it alters the product architecture, the working principle or the business model, so that the designer’s existing competence and the customer’s existing habits are both partly invalidated. It is incremental when the architecture and working principle are retained and the parameters, materials, features or manufacturing route are improved within them.

Radical example 1 — the battery-electric passenger car. The working principle changes from intermittent combustion driving a crankshaft through a multi-ratio gearbox to an electrochemical store driving a motor with a single reduction. The architecture changes with it: no engine block, no exhaust, no fuel system, a structural battery in the floor, a flat interior, and a refuelling model that moves from a five-minute stop at a public station to overnight charging at home. The supply base, the service network and the plant skill set are all displaced. That combination — new working principle, new architecture, new infrastructure — is what makes it radical.

Radical example 2 — additive manufacturing of end-use metal parts. Conventional manufacture removes material from stock or forces it into a die, so cost rises with geometric complexity and falls with volume. Powder-bed fusion builds the part layer by layer, so complexity is nearly free, tooling disappears, and the economic batch size falls to one. That inverts the design rules a mechanical engineer was trained on: internal conformal cooling channels, lattice cores and consolidated assemblies become sensible, while draft angles and parting lines become irrelevant. It is radical precisely because it invalidates design heuristics rather than merely improving a process.

Incremental example 1 — the successive generations of the domestic front-loading washing machine. The architecture has been stable for decades: horizontal drum, suspended tub, single motor, water valve, heater and drain pump. Generation to generation the changes are a direct-drive motor replacing the belt, a variable-speed inverter, load sensing that trims water and cycle time, better door-seal geometry and quieter bearings. Each is a real engineering gain in energy and water consumption, and none of them requires the customer to learn anything new or the maker to abandon its competence.

Incremental example 2 — the bicycle derailleur gaining sprockets. Moving from nine to ten to eleven and twelve sprockets is demanding engineering — thinner chains, tighter chainline tolerance, stiffer cassette bodies, revised ramp timing — but the working principle (deflect the chain sideways onto an adjacent sprocket) and the architecture are untouched. The user’s mental model is unchanged. Contrast this with the electronic derailleur, which is a step closer to radical because it replaces the mechanical cable transmission with a bus, a battery and firmware, and thereby opens up automatic shifting the mechanical version could not offer.

B — How consumer adoption differs (4 marks)

Radical and incremental changes diffuse through a market in visibly different shapes, and the Bass diffusion model captures the difference with two parameters. The cumulative adopted fraction is

where p is the coefficient of innovation — the propensity to adopt from advertising or independent judgement — and q is the coefficient of imitation, the pull of word of mouth and observed use. A radical change has a low p (few people will buy on their own judgement, because the benefit is unproven and the existing solution works) and often a high q (once neighbours own one, the social proof is powerful). An incremental change has the reverse: a high p, because it slots into an existing purchase decision with no learning required, and a moderate q, because there is little to talk about.

Given. A radical product characterised by $p=0.02$, $q=0.55$ and an incremental one by $p=0.14$, $q=0.30$.

Find. The time to peak adoption rate and the cumulative penetration at two and five years for each, so the difference can be stated quantitatively rather than as an impression.

1. Locate the peak of the adoption rate. Differentiating $F(t)$ and setting the second derivative to zero gives the inflection time and the peak rate as$$t^{*}=\frac{\ln(q/p)}{p+q},\qquad \left.\frac{dF}{dt}\right|_{\max}=\frac{(p+q)^2}{4q}$$For the radical product $t^{*}=\ln(27.5)/0.57=5.81$ years at a peak rate of 14.8 % of the market per year; for the incremental one $t^{*}=\ln(2.143)/0.44=1.73$ years at 16.1 % per year.
2. Evaluate the early penetration, where the two differ most. Substituting into $F(t)$ at $t=2$ years,$$F_{\text{rad}}(2)=0.069,\qquad F_{\text{inc}}(2)=0.310$$and by five years the gap has narrowed to 0.364 against 0.719. The radical product therefore spends$$\boxed{t^{*}_{\text{rad}}=5.81\ \text{yr vs}\ t^{*}_{\text{inc}}=1.73\ \text{yr}}$$more than three times as long reaching its peak, and at two years has captured barely a fifth of the share the incremental change has.
3. Read the managerial consequence. The radical curve is not a worse curve, it is a later one: its peak rate is only slightly lower, and its high imitation coefficient means it accelerates hard once it starts. What it demands is a cash runway long enough to survive four to six lean years, a tolerance for early adopters who behave nothing like the mass market, and investment in the complements — charging networks, service training, standards — that reduce the adopter’s risk and so raise $p$.

Qualitatively the same conclusion follows from Rogers’ five attributes of an innovation. Incremental change scores well on compatibility, low complexity, trialability and observability, and its relative advantage is small but certain. Radical change offers a large relative advantage but scores badly on the other four, and those four are what govern the speed of the early market. Adoption is limited by perceived risk, not by measured benefit.

C — What makes a society more open to radical change (3 marks)

Openness to radical change is largely a matter of how cheaply a society lets an individual or a firm be wrong. The factors that matter most:

Tolerance of failure and the cost of trying. Where a failed venture is a survivable episode rather than a permanent stigma — workable bankruptcy law, portable health coverage and pensions, an active venture-capital market, employment rules that let people move — more people will attempt the unproven. Canada’s Scientific Research and Experimental Development tax credit works on exactly this margin by lowering the private cost of an experiment.

Wealth, education and demographic profile. Discretionary income lets a household absorb the risk of an unproven product; technical education makes the benefit legible; and a younger, urban, digitally connected population has both lower switching costs and denser word-of-mouth networks, which is the $q$ term above.

Institutions that reduce the adopter’s risk. Enabling infrastructure and public procurement (charging networks, broadband, a government fleet buying first), standards and certification that make quality verifiable, and intellectual-property protection that makes the innovator willing to disclose. Conversely, prescriptive rather than performance-based regulation, entrenched incumbents with regulatory capture, and a strongly uncertainty-avoiding culture all slow radical adoption — and a visible shock, whether an oil-price spike or a pandemic, can suspend that resistance almost overnight.