Dematerialisation of products and manufacturing-generated knowledge content: relationship through paradigms

Abstract

Dematerialisation refers to the decline, over time, of material usage in industrial products. The subject of this paper is to identify the relationship between the material and knowledge content of several products. The issue is whether the knowledge content, through time, will be affecting the products, in terms of dematerialisation. Specifically, a study has been conducted on widely used products in order to determine the way they have changed through their life-timeline. Two different indicators have been introduced for the quantification of the trend, in terms of material and knowledge content.

Keywords:

• knowledge content
• dematerialisation
• knowledge measurement
• sustainability
• lean manufacturing

1. Introduction

During the last few decades, the manufacturing world has been changing dramatically; an indicative example, regarding manufacturing-related criteria (Chryssolouris 2006), is the integration of sustainability-related aspects (Chryssolouris, Papakostas, and Mavrikios 2008). There is a trend for products to utilise, in general, less material and yet perform even better. This is the dematerialisation trend and its definition may be (UNTERM 2016): Dematerialisation, ultimately describes the decrease in material requirements of whole economies, comprising:

(a)	The reduction in the material-intensity of products and services
(b)	The reduction in the use of primary material resources by improving the recycling and re-using of secondary materials

There is a connection between dematerialisation and material consumption. There, the issues are more complex, because the products themselves may have been made of less material and yet, the consumption of all the materials to remain high, due to the higher consumption of these products. Both the rapid growth of population globally (UNEP 2015) and the globalisation itself contribute to that increasing consumption. The total material consumption per capita (Figure 1(a): Total Material Consumption per capita) remains steady or follows a slight increase in some geographical areas (OECD members). The case of the BRICS members though, is that the total material consumption per capita, follows a steady increase. As the population keeps growing, in combination with the consumers’ increasing requirements, the absolute global material consumption (Figure 1(b): Total Material Requirements), does not decline, as it would have been the case of environmental sustainability.

Figure 1. Total material consumption (a) and requirements (b) (KNOEMA 2016).

There is an issue, related to the effect of a product on the environment (UNEP 2016), because a product may have been made of less material and yet, have a greater impact on the environment if the material is not recyclable, it has a short lifespan etc.

The impact of the dematerialisation, as an environmental sustainability measure, is a major issue on related studies, along with suggestions that have been made for gauging dematerialisation and any additional actions towards it.

A quantitative analysis of dematerialisation has been proposed by Magee and Devezas (2014). This study introduces a dematerialisation criterion that includes technical performance changes, through time and demands a rebound effect. A great analysis on the dematerialisation trend and the way that the world has come to its current state, concerning energy and material demands, has been made by Smil (2013).

Dematerialisation is based on the improvement of the products’ efficiency and on reusing or recycling materials and products. An ideal framework of dematerialisation would concern any action taken in every stage of the production and the consumption line. Those actions may refer to resource savings in material extraction, an improved, more efficient design of the product, the use of technological and engineering innovations in the production stages and even the consumers’ environmental consciousness, with the use of eco-friendly products and recycling. Therefore, from an environmental point of view, dematerialisation could also be defined as any change in the amount of waste generated per unit of industrial products (Wernick, Govind, and Ausubel 1996).

An additional aspect of dematerialisation could be the so-called ‘servitization’ of products. Product-Service Systems (PSS) contribute on strategies towards dematerialisation and thus, towards sustainability (Wang et al. 2011; Durugbo 2013; Mahut et al. 2017). In (Belvedere, Grando, and Bielli 2013) a study related to Information and Communication Technologies, discusses how ICT can contribute to value creation of a company.

Herman, Ardekani, and Ausubel (1988) first questioned whether dematerialisation was actually happening or not. A serious discussion, considering the quality of any dematerialised products and suggesting that dematerialisation on its own, does not contribute to a successful environmental policy, has been performed on that project.

There is an issue of the knowledge content and the way that one can estimate one’s knowledge about a product. Several researchers have studied the subject of knowledge and its valuation (Bernard and Xu 2009; Xu and Bernard 2010; Xu et al. 2014; Magee et al. 2016). Knowledge representation and knowledge creation in manufacturing systems and in product development in general, is presented in (Henriksen and Rolstadås 2010; Wu et al. 2014). Additionally, others researchers (ElMaraghy and Urbanic 2003; ElMaraghy, Kuzgunkaya, and Urbanic 2005; Chryssolouris, Vasiliou, and Mavrikios 2006; Lu and Suh 2009; ElMaraghy et al. 2012; Chryssolouris et al. 2013; Smart, Calinescu, and Huaccho Huatuco 2013) used the information theory, complexity measures and uncertainty in an attempt to quantify the information exchange and thus, the knowledge content, throughout the production of a certain product. A general definition of knowledge is:

•	Information, understanding or skill that you acquire from experience and/or education
•	awareness of something

The main question to be discussed in the present paper is if dematerialisation is also accompanied by an increase in knowledge. If such a correlation exists, it might lead to a better understanding of the industry’s state regarding sustainability. Innovative products, accompanied with a factor of dematerialisation, would immediately imply an increase of knowledge. As dictated by modern vision (Cho, Alamoudi, and Asfour 2009; Toro, Barandiaran, and Posada 2015) and predicted by (Chryssolouris 2006), information exchange takes place throughout the product development system, and with the use of methods and models, based on the Information Theory, this exchange becomes quantifiable and measurable. On this basis, and the assumption that the functionalities of the final product represent, in a convincing way, the embedded information, through the study of a product’s historical advancement, one can estimate a possible knowledge increase. Additionally, metrics such as the mutual information of two variables, in this case, the product’s functionalities, can be a measure of knowledge related to the product. As such, one could establish a framework in the information embedment efficiency and thus, a measure in the sustainability (dematerialisation), as well as the optimisation efficiency.

The subject of this paper is to establish a preliminary framework for examining any potential relationship between a product’s dematerialisation trend and its knowledge content. Namely, if products utilise less and less material, there will also be an increase in their knowledge content.

2. Approach

A notable transition, from matter to information, almost happens to every product, with the implication that the information becomes more valued than the matter does. More information and knowledge are inherent in every new product; for instance, first generation mobile phones, compared with the last generation smartphones, or the CRT computer monitors compared with the modern LED displays. Despite this transition, the ‘global’ consumption of resources keeps increasing, due to the consumption of advanced products by the growing world population. This phenomenon is known as Jevon’s Paradox (Alcott 2005): since the technological progress increases the resource’s efficiency, its rate of consumption rises, due to the increasing demand, caused by that improved efficiency. This paradox describes the current state for a number of ‘technological’ products.

From a manufacturing point of view, the dematerialisation of products is affected by a number of factors such as: ease of manufacturing, production cost, quality, size and complexity of the product (Herman, Ardekani, and Ausubel 1988). An easily manufactured, high-quality product, will save time, energy, waste material and thus money. This is a step forward, in the direction of dematerialisation and environmental sustainability.

The motto ‘produce more with less’ (Chryssolouris, Papakostas, and Mavrikios 2008), describes ideally the dematerialisation approach as to the way new products are produced in comparison to those in the past.

In order for the dematerialisation and knowledge content to be discussed, two indicators are introduced in the context of this study: the material and the knowledge indicators. These two indicators somewhat describe the dematerialisation trend and the knowledge content and constitute a general framework to address material use vs knowledge content.

This work involves in a rather simple way, the study of some relevant products, regarding the evolution of material use and knowledge content over time, in order to apply the general framework (Figure 2) in specific cases.

Figure 2. Flowchart of the framework application.

3. Results

A large number of products get lighter and smaller through their evolution. A relative form of dematerialisation happens to many modern products. In order for any relationship, between material content (dematerialisation) and knowledge content to be discovered, there has been an examination undertaken of several products that enable, over time, the determination of the value of these indicators and any relationship between the two.

In each case, the primary criterion for the definition of knowledge has been the advancement of the product’s technical performance through time. Thus, it is simply stated that for any improvement in a product’s performance and/or efficiency, an amount of knowledge has to be considered in order for this outcome to be possible.

An indicator K regards the normalised value of knowledge as it arises from the consideration of knowledge on each product, respectively, and an indicator M regards the normalised value of the material used (see Table 1).

Figure 3. Material – Knowledge correlation.

Table 1. Indicators M and K Interpretation for each product.

Product	Indicators interpretation
	Material indicator – M	Knowledge indicator – K
Storage disc	Disc Weigh/Price	Capacity/Price
Laptop	Laptop Weight/Price	Processor Speed/Price
Car engine	Engine Net Weight	Power/Weight
Light bulb	Net Weight needed for 50,000 h	Lumens/Watt
Refrigerator	Weight/Capacity	Energy Consumption/Capacity
Electronics	Products Net Weight	Transistors/mm^2
Material	–	Energy Intensity ‘e’

The results produced a correlation between material and knowledge, as presented  in Figure 3:

In the Appendix 1, the numerical values of the indicators K and M are presented along with the graphic representation of the results for each case.

4. Discussion

As it appears from a preliminary analysis, there is a strong trend of product dematerialisation. However, as the efficiency of a product increases, the demand for that product may also rise, consequently, the overall material consumption may rise despite the decrease in material usage of the individual product. Between dematerialisation (M) and knowledge (K), one could formulate a form of a conservation law such as:

The preliminary analysis of products, has proven in a rather simple way that dematerialisation happens for all of the selected products.

Using the regression analysis on each product, the following results have been derived by assuming that,

with t being time and λ defined as the ratio B/A (see Table 2).

Table 2. Embedded knowledge and material intensity on various products: exponents-based analysis.

	A-exponent (material)	B-exponent (knowledge)	λ = B/A	Product character
Hard drives	1.8	10.9	6.05	λ > 1
Laptops	1.2	5.7	4.75	λ > 1
Cars	2.5	1.2	0.48	λ < 1
Light bulbs	2.18	2.23	1.02	λ = 1
Refrigerators	1.22	1.25	1.02	λ = 1
Printers	0.38	6.5	17.1	λ >> 1
Material	N/D	2.12	Defined as 1	N/A

Conclusions, regarding the products’ knowledge intensity, can be derived from the above. For a high value of λ, the knowledge intensity is higher compared to that of a low value. For example, hard drives and laptops (i.e. electronics) appeared to be more knowledge intensive than cars.

5. Conclusions

In conclusion, this first simplified approach of knowledge value to dematerialised products indicates that a relative form of dematerialisation, concerning commonly used products, has taken place. An increasing rate of the K value is observed on all the products that have been examined, and a declining rate of material use on those products is also observed. Electronics devices, including laptops and hard drives, appear to be the most dematerialised products.

There are some limitations regarding the immediate use of the proposed concept. Specifically, the consideration of the knowledge content on each product, may differ, depending on one’s understanding of this knowledge. As suggested in Section 1, an enhanced framework on the information embedment efficiency, via metrics related to the functionalities of products, could be established in an attempt to measure both their sustainability (dematerialisation) and optimisation efficiency. To this effect, the immediate steps to be made, link the product-embedded knowledge with manufacturing-related information and furthermore, dematerialisation with sustainability.

Nomenclature
M	=	Material indicator
K	=	Knowledge indicator
A	=	Material indicator evolution exponent
B	=	Knowledge indicator evolution exponent
C1	=	Value of material indicator for time equal to zero
C2	=	Value of knowledge indicator for time equal to zero
t	=	Time
λ	=	Product character indicator

Disclosure statement

No potential conflict of interest was reported by the authors.

Appendix 1

Paradigm A: Storage Discs Paradigm

The following assumptions have been made for this analysis:

•	Storage capacity is considered being the most important technical performance aspect of a storage disc
•	The selected products are considered being a State-of-the-art product from their first appearance
•	Data retrieved from various sources: (Average Cost of Hard Drive storage 2016; IBM 2016; SEAGATE 2016)

Paradigm B: Laptops Paradigm

•	The processor’s speed is considered being the most important aspect of a laptop’s technical performance
•	Data retrieved from various sources: (APPLE INC 2016; Laptops Data 2016)

Paradigm C: Car engines Paradigm

•	Engine of the same manufacturer (Ford)
•	Power to weight is generally considered being an engine’s performance indicator
•	Data retrieved from: (Ford Classics 2016; Ford Engines 2016)

Paradigm D: Light Bulbs Paradigm

The present analysis is based on a technology comparison. Specifically, a comparison has been made between incandescent, CFL and LED technologies of light bulbs.

•	Comparison of light bulbs, producing the same amount of lumens (800)
•	Data Retrieved from: (Lumen Output: Comparing LED, CFL Incandescent Wattage 2014)

Paradigm E: Refrigerator Paradigm

•	The analysis has been made on refrigerators of a similar size and of the same brand (~400 lt/General Electric). Official weight data were not available prior 1995.
•	Data Retrieved from: (General Electric Refrigerators 2016)

Paradigm F: Electronics Paradigm

In this paradigm, there is a presentation of two electronics devices, showing the way they become lighter as time goes by. This paradigm is based on Moore’s law; knowledge consideration will be the definition of the law.

•	The products presented are printers and cell phones
•	The printers selected are of the same manufacturer
•	The selected processors are of the same manufacturer and a criterion for their selection was the reduction in the manufacturing scale process for every new model
•	The process could be an alternative metric of knowledge (k)
•	Smartphones have not been considered, only cell phones
•	Data for processors retrieved from: (Microprocessor Quick Reference Guide 2008)
•	Data for Cell phones retrieved from various sources
•	Data for printers retrieved from: (HP Deskjet Printers 1988–8)

Paradigm G: Material Paradigm

In this paradigm, the material energy intensity ‘e’ measure is used. As defined in (Gutowski et al. 2013), energy intensity is: ‘the energy required to produce a material from its raw form per unit mass of the material produced’^{Figure 4Table 3}

•	As the value of ‘e’ decreases, one can assume that the knowledge value increases
•	Indicator K, states that for the same amount of produced material (be it a ton of steel) the energy requirements have been declining through time. This decline implies that there is a relative decrease in the fuels, required for the production of that energy, whether that energy is by fossil fuels (coal, coke etc.) or the energy of an electric arc furnace (electricity), etc.
•	The indicator M at this paradigm is irrelevant
•	Data retrieved from: (US Energy Requirements for Aluminium 2007; World Steel Association fact Sheet 2016) (see Figure 4 and Table 3)

Figure 4. Dematerialisation – Knowledge correlation: Products paradigms.

Table 3. Dematerialisation – Knowledge correlation: Products paradigms.

A. Storage discs
Model	Year	Weight (kg)	Storage capacity	Price ($)	m (weight/price)	k (Capacity/price)	Μ (normalised m)	Κ (normalised k)
IBM	1956	1000	5	50,000	0.02	0.0001	1.0000	0.0000
Pegasus great lakes	1984	13	23	1845	0.0070461	0.012466125	0.3034	0.0000
Seagate st-251	1989	1.25	40	490	0.002551	0.081632653	0.0617	0.0000
Quantum 1080S	1995	0.9	1050	430	0.002093	2	0.0371	0.0001
WD – WDAC33100	1997	0.49	3100	349	0.001404	9	0.0000	0.0003
Quantum 19 GB Bigfoot	1999	1.24	19,000	359	0.003454	53	0.1102	0.0016
Maxtor U-ATA	2001	0.6	60,000	329	0.0018237	182	0.0226	0.0057
Hitachi Deskstar	2007	0.7	1,000,000	370	0.0018919	2703	0.0262	0.0838
WD – Caviar	2010	0.73	2,000,000	200	0.00365	10,000	0.1208	0.3100
Seagate Backup Plus 4 TB	2013	0.247	4,000,000	124	0.0019919	32,258	0.0316	1.0000
B. Laptops
Model	Year	Weight (kg)	Processor speed (MHz)	Price ($)	m (weight/price)	k (processor speed/price)	Μ (normalised m)	Κ (normalised k)
Osborn 1	1981	10	4	1795	0.0056	0.0022	1.0000	0.0003
IBM 5155	1984	13.6	4.77	4225	0.0032	0.0011	0.5451	0.0000
SLT/286	1988	6.4	12	5399	0.0012	0.0022	0.1519	0.0003
NEC Ultralite	1989	2	8.14	5000	0.0004	0.0016	0.0000	0.0002
PowerBook 100	1991	3.63	16	2300	0.0016	0.0070	0.2279	0.0018
ThinkPad 700	1992	2.67	25	2950	0.0009	0.0085	0.0977	0.0023
PowerBook 5300	1995	2.7	100	2300	0.0012	0.0435	0.1497	0.0130
ThinkPad 380	1997	3.19	150	2719	0.0012	0.0552	0.1495	0.0166
PowerBook G4	2001	2.4	400	3499	0.0007	0.1143	0.0553	0.0347
Vaio TX6708	2005	1.2	1200	2199	0.0005	0.5457	0.0282	0.1670
MacBook Air	2008	1.36	3200	1799	0.0008	1.7788	0.0688	0.5452
MacBook Air	2010	1.35	3400	1599	0.0008	2.1263	0.0859	0.6518
MacBook Air	2015	1.35	4400	1349	0.0010	3.2617	0.1162	1.0000
C. Car Engines
Ford Engine	Year	Weight (kg)	Power HP (Watt eq.)		m (engine net weight)	k (power/weight)	Μ (normalised m)	Κ (normalised k)
Model T	1921	230	20 (14,710)		230	0.09	1.0000	0.0000
Crossflow	1977	117	63 (46,336)		117	0.54	0.1630	0.3839
Zetec Engine	1992	107	89 (65,459)		107	0.83	0.0889	0.6332
Zetec SE	2002	97	74 (54,427)		97	0.76	0.0148	0.5747
Ford EcoBoost	2015	95	120 (88,260)		95	1.26	0.0000	1.0000
D. Light Bulbs
						Incandescent	CFL	LED
Weight (grams)						31	100	185
Lifespan (hours)						1200	10,000	50,000
Light bulbs needed for 50,000 h						42	5	1
Watts for 800 lumens						60	15	8
Compare light bulbs that produce 800 lumens
m (overall weight in grams)						1302	500	185
k (lumen/watt)						13.33	53.33	100
K (normalised k)						0	0.46	1
M (normalised m)						1	0.28	0
E. Refrigerators
Model number	Year	Energy consumption	Capacity (lt)	Weight (kg)	m (weight/capacity)	k (consumption/capacity)	M (normalised m)	K (normalised k)
TBF14BV	1979	119 kWh/month	402	–	–	0.2815126	-	0
TBF15DB	1984	107 kWh/month	424	–	–	0.3302181	-	0.070603
TB15SG	1989	65 kWh/month	412	–	–	0.5282051	-	0.357602
TBH14DAS	1995	42 kWh/month	407	83	0.2039312	0.8075397	1	0.76252
TBX12SYB	1999	45 kwh/month	408	76	0.1867322	0.7555556	0.695802	0.684481
GTS15BBP	2004	37.5 kWh/month	422	66.67	0.1579858	0.9440716	0.187366	0.960435
A3316ABS	2009	38 kWh/month	441	65	0.1473923	0.9713656	0	1
F. Electronics devices
Knowledge consideration – Indicator K
Microprocessor			Year	Transistors	Size mm^2	Process (nm)	k (size/transistors)	K (normalised k)
intel 4004			1971	2300	12	10,000	191.6666667	0
intel 8008			1974	4500	14	10,000	321.4285714	8.52E-05
intel 8088			1979	29,000	33	3000	878.7878788	0.000451
intel 80,386			1985	27,500	104	1500	264.4230769	4.78E-05
intel 80,486			1989	1,180,235	175	1000	6744.2	0.004304
Pentium			1993	3,100,000	294	800	10,544.21769	0.0068
Pentium II Deschutes			1998	7,500,000	113	250	66,371.68142	0.043473
Pentium 4 prescott			2004	112,000,000	110	90	1,018,181.818	0.668705
core 2 duo allendale			2007	169,000,000	111	64	1,522,522.523	1
Material consideration – Indicator M
Cell phone paradigm						Year	m (product net weight – kg)	M (normalised m)
Motorolla DynaTex 8000						1983	0.794	1
Nokia Cityman						1987	0.760	0.952381
Motorolla International 3200						1992	0.576	0.694678
Motorolla Microtac						1994	0.350	0.378151
Ericsson GH688						1996	0.160	0.112045
Motorolla Timeport						1999	0.140	0.084034
Nokia 3510i						2002	0.106	0.036415
Samsung E250						2005	0.08	0
Printers paradigm						Year	m (product net weight – kg)	M (normalised m)
Deskjet						1988	6.5	0.977778
Deskjet 500						1990	6.5	0.977778
Deskjet 550C						1992	6.6	1
Deskjet 560C						1994	6.6	1
Deskjet 680C						1996	5.3	0.711111
Deskjet 895						1998	5.5	0.755556
Deskjet 648C						2000	5.26	0.702222
Deskjet 3420						2002	3	0.2
Deskjet 469						2005	2.1	0
G. Materials
US steel industry energy intensity
Year						Energy use (kWh/kg)	k = 1/(kWh/kg)	K
1950						14.5375	0.068787618	0
1960						13.2453	0.075498479	0.048779
1970						11.953	0.083661006	0.10811
1980						9.6917	0.103181072	0.249995
1990						6.4611	0.154772407	0.624995
2000						5.4919	0.182086345	0.823532
2010						4.8458	0.206364274	1
US aluminium industry smelting energy
Year						Energy (kWh/kg)	k = 1/(kWh/kg)	K
1960						23.1	0.04329	0
1970						21.4	0.046729	0.149942
1980						17.5	0.057143	0.604
1990						16.1	0.062112	0.820652
2000						15.1	0.066225	1