The D. Allan Bromley
Memorial Lecture:
Bringing Advanced
Innovation to
Manufacturing
William B. Bonvillian
Director, MIT Washington Office
University of Ottawa,
May 14, 2012
D. Allan Bromley
2
Background:
The U.S.-Canada Partnership
 U.S.-Canada – the largest bilateral trade relationship
in
the history of the world –
 Total merchandise trade: $530B (2007 pre-recession)
 Enhanced through deep trade agreements - U.S.Canada Auto Pact of 1965 to NAFTA
 U.S.-Canada economies highly integrated – comparable
standards of living and industrial structure
 U.S. (pre-recession 2007): $300B in in imports from Canada
and $230B in exports
 We must care a lot about each other
 I bring today a perspective on U.S. manufacturing but it
affects both nations
 I will outline steps I believe the U.S. must consider, and have
been active in a number, but will leave details on the
actions and who the actors might be to Q&A
3
4
BACKGROUND POINTS
---why “manufacturing
matters”
Hollowing Out?
5
Employment:
 Down almost 1/3 in a decade
Investment:
 Manufacturing fixed capital investment declined
(accounting for costs) in the 2000s for the first time since
the data has been collected
Output:
 Adjusting gov’t data (for foreign component origin and
inflationary assumptions in IT and energy sectors),
manufacturing output value declined in the 2000s
 Decline in 16 of 19 sectors
Productivity:
 If output lower than assumed, productivity is lower
We have been assuming we
have been losing
manufacturing jobs because of
productivity gains
 Recent work – “The Race Against the Machine,” for example is telling us that multiplying productivity gains from IT are
displacing work as we know it.
 Yet, historically - most recently in the tech boom of the 90’s productivity gains, although disruptive initially, grow more jobs
 Maybe that history is still true – and maybe we need to search
for our profound job losses in the manufacturing sector.
 That means “The Great Recession” is structural, not business
cycle, so the Keynesian stimulus tools we have been applying
won’t work well.
 The Manufacturing Hollowing Out is why these aren’t working
– requires a Structural strategy not a maco-economic strategy
6
Sharp Decline in Mfg. Employment,
2000-2010 -- drop so steep that
productivity gain can’t explain
7
Manufacturing Remains a
Major Sector
 Manufacturing = $1.7 Trillion of $15T U.S. economy
 Employs 12 million in workforce of 140m
 Dominates the U.S. innovation system – 70% of of industrial
R&D, 80% of patents, employs 64% of scientists and engineers,
 The currency of international trade is complex high value
goods –
 80% of U.S. exports are high value goods (capital goods,
industrial supplies, transport goods, medicines)
 U.S. pre recession ‘07: $500B deficit in goods – on track to
return to that level
 Services surplus ($160B) growing gradually but will not offset
manufacturing deficit in foreseeable future
8
9
Underlying Issue: Our
“Innovate here/Produce
Here” Assumption
 Since WWII - U.S. economy organized around
leading the world in technology advance.
 US led all but one of the innovation waves of the 20th
century – and growth economics tell us that
technological & related innovation = 60%+ of growth
 from aviation to electronics, to nuclear power, to
computing, to the internet, to biotech
 Missing an innovation wave is serious: Japan led quality
mfg.; 1973-1991 tough for U.S. – GDP and productivity 1%
below historical averages
 Response: ‘90s IT innovation wave and record growth
 Our operating assumption - we would innovate here and
WE would translate those innovations into products
 Would realize the full range of economic gains from
innovation at all stages
 It worked – world’s richest economy
10
“Innovate here/Produce
here” Bonds Breaking?
 With global economy, assumption of innovate
here/produce here no longer holds.
 In some industrial sectors, can now sever R&D and
design from production –
 That brings the economic foundation of our
innovation-based economy into question. Why
invest in innovation here if gains elsewhere?
 Last 25 years – IT/electronics allowed severing of
R&D/design from production via IT-based specs;
commodity goods, too
 Distributed Manufacturing – Apple iPod example
 But other sectors still require deep connection
between R&D and production – constant
reengineering and improvements to cut costs
11
Mind The Connection between R&D/Design and
Production in Different Sectors
12
 3 basic kinds of produced goods:
 (1) IT, (2)bio/pharma, and complex (3) electromechanical-aero
 First – can sever R&D/design from production using batch
processing (bio/pharma), IT specs (IT goods)
 Electro-mechanical-aero – tie R&D/prod.– variables too
complex
 Energy, for ex., is in the 3rd electro-mechanical category –
need to connect R&D/design with production
 If offshore production, will design/R&D follow?
 Distributed Mfg.: risk losing production;
 the rest: lose production, will design/innovation follow?
 Underlying all this: Competing with low cost/wage high tech
competitors: must have production productivity gains
 That means new innovation req’d: technology and processes
If Manufacturing is changing,
what are the New “Geopolitics of
Manufacturing”?
13
 U.S. has gone through 3 phases:
 1789-1945 – Alexander Hamilton saw that U.S. independence
and security would be built on its commercial strength – U.S.
pursued strong self-dependence in manufacturing through
WW2
 1945-1993 – Cold war competition with Marxist economic
system – U.S. strategy: North America, Western Europe, Japan
in a system of mutual economic dependence and integrated
economies among allies.
 1993-now: Clinton: unified global economy is way to integrate
China into the world system – manufacturing would be global –
decentralized and integrated
 If manufacturing is no long a tool of national security, can
technology leadership shift?
Historical Examples of
Manufacturing Shifts
14
Post-WW2, U.S. built a comparative advantage in
innovation – but unlike the Ricardo comparative
advantage in resources, comparative advantage
in innovation can shift – it is not eternal (Samuelson 2004)
1.
US takes leadership of Industrial Revolution mid- 19th
century through development of the “American system” of
interchangeable machine-made parts
 Result of 20-year DOD technology development of
precision machine tools at Harper’s Ferry Arsenal
2. Japan 1970’s-80’s – Quality Mfg. - new quality/price
tradeoff, just in time inventory, making labor a fixed cost for
labor flexibility – tech: IT, computer driven machine tools,
tied to production process - built in quality at every phase
3. US recaptures Semiconductor manufacturing lead in 80’s –
focus on mfg. process – advances in Semiconductor equip.
suppliers, roadmapping
In All Three Shifts: tech innovation, process, business model
Competiveness Then and Now
– much more complex
JAPAN v. U.S. – 1970’S – 80’S
CHINA v. U.S. - Today
High cost, high wage, advanced
technology economy –
comparable to the U.S.
Low cost, low wage, increasingly
advanced technology economy
U.S. had entrepreneurial
advantage, Japan had industrial
policy advantage
Entrepreneurial and pursuing
industrial policy
Rule of Law
Limited Rule of Law
IP Protections
Extensive IP theft
15
Subsidizes currency, buys U.S. debt Following Japan’s model: subsidizes
currency and largest holder of U.S.
debt
This Competition = Eroding U.S.
Advanced Technology Sectors
 Shifted abroad:
 Every brand of notebook computer
 Every mobile/handheld
 Displays
 Shifting abroad? Major erosion in:
 Advanced materials
 Computing and communications
 Renewable energy technologies and storage
 Semiconductor production
 The Kindle could not be made in the U.S.
 $100B trade deficit advanced tech goods
16
Behind it all:
Understanding
the Hourglass -<---- Resources, Suppliers,
Components,
Innovation
<--- Production (12m jobs)
<--- Distribution, Sales, Life
Cycle
17
It is an Hourglass because
we know Mfg. is a Job
Multiplier:
 Economic Policy Institute: Mfg. job
multiplier - 2.90
 Milken Institute: 2.50 multiplier
 Zobel (Germany) smart manufacturing
factory supports 5.2 additional jobs
 High-tech manufacturing industries
appear to have greater multipliers Electronic computer manufacturing
multiplier of 16
 [ITIF 4/11 summary]
18
Change in Median Household
Income, 2000-2011 (inflation adj.)
+2
0
-2
-4
-6
-8
-10
-12
CAUSE: STRUCTURAL RECESSION – Centered in Mfg.
Source: G.Green, J.Coder (10/11, based on Census Bur. Data)
19
If it is a structural
recession with a base in
manufacturing, this
requires:
Innovation in
Advanced
Manufacturing:
9 Steps to
consider
20
Need to Keep in Mind the
Levels in the US
Manufacturing Sector
21
 Three elements in US mfg.:
 Large Multinationals (MNCs) – very international - will
locate in low cost production centers for productivity
gains; need to be in emerging markets
 Start-up and entrepreneurial firms – increasingly
offshoring production – not a core competency, can’t
get financing So: next gen technologies shift?
 300,000 small manufacturers – suppliers, component
makers – bulk of US mfg
 Thinly capitalized, risk adverse, no R&D
Step #1: What technology advances
= new manufacturing productivity
= new paradigms?
22
 “Network centric”
– mix of advanced IT, RFID, sensors in every stage and element, new decision making from
“big data” analytics, advanced robotics, supercomputing w/adv’d simulation & modeling
 Advanced materials

“materials genome” – ability with supercomputing to design all possible materials with
designer features
 Biomaterials, bio fabrication, syntehtic biology
 Lightweighting everything
 Nanomanufacturing
 fabrication at the nano-scale
 Mass Customization
 Production of one at cost of mass production (ex.: 3D printing/additive mfg, etc.)
 Distribution efficiency
 IT advances that yield distribution efficiencies (incl. in supply chain)
 Energy Efficiency – energy is “waste
These tech paradigms are
not optional – Companies
will have to meet them
 Timetable – starting to see these adv’d mfg.
technologies emerge – will be pervasive in next 15 years
 Just as with “Lean Mfg.” (way US firms responded to
Japan’s model of the 70s-80s), today’s firms will need to
pursue these advanced technologies because the
competitive efficiencies are major
 How will mass of 300,000 US SMEs adapt to these tech
advances? Big challenge
 US: strength on the design side – real limits on the
commercialization and scale-ability side which
advanced mfg. innovation would enable
 US is not yet working systematically on this agenda
23
Step #2 – Pick Tech Paradigms
that Apply across Sectors
Manufacturing is sectoral, but with
increasing sectoral overlap for
complex, high value goods
An airplane is complex system: aero
design, electronics, IT, materials, etc.
Technology paradigms have to make
sense in the sectors
Run a matrix – technology options
against sectors they apply to – pick
technologies with payoff across
sectors
Include emerging sectors
24
MATRIX: Tech Sectors/Mfg. Paradigms
25
Sector
and Mfg.
Pardigm
Bio/pha Aerorma
space
IT/elect
ronics
Heavy
Equip
ment
Digital
search,
network
New
energy
Trans
port
Network centric
x
x
x
x
x
x
x
Advanced
materials
x
x
x
x
x
x
Nano Mfg.
x
x
x
x
x
x
x
Mass
Customization
x
x
x
x
x
x
x
Distribution x
Efficiency
x
x
x
x
x
Energy
Efficiency
x
x
x
x
x
x
x
Step #3 – Technology is Not
Enough…Need to Look at
all Three:
A) New Adv’d Mfg.
Technology Paradigms - first
building block
B) But need Process – adapt to
production system
C) Then need Business and
Organizational Model – has to
26
Step #4:, It’s no longer Manufacturing
or Services
 21st Century firm increasingly fuses services,
production, supply chain management and
innovation –
 IBM’s Lou Gerstner originated this model in the 90’s
 Many of these capabilities are knowledge
“intangibles” not fixed assets – learning to tie new
equipment and technologies to new processes –
fusing IT-informed services models with new mfg.
 What is this emerging firm model? How pervasive?
 is it vertical or horizontal?
 is it integrated or the result of flexible leveraging
other firms’ specialty capabilities?
 Business model stage - will need to look at optimal
combined mfg./services model
27
Step #5: Need to Look Over
Our Shoulders
28
Look at competitor nation strategies
Hard to understand the future of U.S.
manufacturing without evaluating the context
of global manufacturer competitors and their
strategies – learn from them – they are doing
this and have “top down” not just “bottom up”
Look at:
China/India/Brazil – large emerging
Germany/Japan – large established
High wage & cost – yet major mfg. surpluses
Korea/Taiwan – smaller scale, key niches
Expanding mfg. employment
Step #6: Build the Workforce  Most of mfg. is now defined as services –
 Need a new level of fused knowledge, skills in both
29
 STEM Ed req’d; mfg.: pervasive basic STEM skill sets
 Ed system doesn’t understand that innovation requires “mind and
hand”
 “intelligent hand” - mix of skills, experimentalists and theorists –
learning by doing
 Additive manufacturing in schools?
 It’s not just design as a stand-alone stage, design is over time
also the ability to make, as well – education needs to incorporate
 Very hard, still, despite distributed IT manufacturing, to sever
design from production – mutually informative
 Community college skills role w/industry certification
 New adv’d mfg. engineering curriculum – how to develop
process from lab bench technologies – lost in engineering
curriculum
Step #7: Innovation Organization The Pipeline and the Seams
30
 US pipeline innovation model organized with heavy
federal basic research investment,
 some applied (from DOD)
 Limited investment in manufacturing R&D (including tech, process,
business model) - $800m – and not interagency
 We institutionalize the “Valley of Death” in our R&D model
 profound problems at the seams of the innovation pipeline – big
disconnects between actors
 U.S. research strong; scaling/commercialization a problem:
 Research – basic research agencies, univ’s
 Applied and later stages – industry, some DOD support
 But other intermediate steps will need public-private
connections –further down the pipeline to commercialization
 Not just R&D–pre-production networked organizational
We Need these Guys - DOD: Operates
at Every Stage of the Innovation System Historically Central Player
- Role in Adv’d Mfg.?
REMEMBER DOD’s 20th Century
Innovation Waves:
 Aviation
 Electronics
 Nuclear Power
 Space
 Computing
 The Internet
31
Step #8: Build Regional
Infrastructure
32
Research > Development > Prototype >
Demonstration > Testbed > Production at Scale
 No longer the Old Pipeline model – gov’t does Research
and industry picks up the rest
 Now: host of intermediary steps require public/private
connections
 Especially important: the Testbed
 Where 300,000 SME’s test the efficiency and cost of new
mfg. technologies – they lack resources for this stage
 Mfg. is regional not national – need to be in regional
clusters
Step #9: Financing Innovation - “5
Year Yardstick” doesn’t work in Mfg.
33
 The Breakthrough system – the pipeline: federal R&D,
univ. research, startups/entrepreneurs, VC’s angel,
IPO’s
 5-year yardstick based on IT model: VC’s fund
technologies no more than 2/3 years from
commercialization, then the flip to an IPO within 3
years
 New manufacturing technology paradigms probably
require the breakthrough innovation system
 But advanced manufacturing doesn’t fit the 5 year
yardstick
5-Year vs. 10-year Yardstick:
3
4
 Manufacturing doesn’t fit the 5-year yardstick:
 New mfg. tech’s face the Valley of Death –
 Then they face the “Mountain of Death” – getting to market launch at
scale: major financing and price competitive at the outset of launch
+
Valley of Death – 5 year scale up
- creating connections and
funding to move from research
to late stage development
Mountain of Death: 10+ year
scale up; major financing needed
to scale, price competitive from
moment of market launch
Work-Arounds for the
Mountain
35
 Manufacturing – the 10+year yardstick
 Requires deeper, longer term, patient capitalization/finance
than IT
 Longer time to stage entry and to scale – 10+ years not 5
 It’s a complex, established “legacy” sector
 US better at bringing innovation into new areas, not at introducing
innovation into legacy areas
 Different mindset – can’t create a company to sell it, as in IT, biotech
 Work-Arounds for the Mountain of Death:
 Front end of the innovation system:
 R&D programs designed for the breakthrough and to drive down
prices/reduce production costs
 US research agencies: “NMP” – Not My Problem – look at just research
not the implementation – change?
 ARPA-E and EERE considering
 Back end of the innovation system?
 Andrew Lo – portfolio approach
 Small commitments by large nos. of investors – use the internet
 Crowd-funding? Lotteries?
Remember Innovation
Organization Effects for the 3
kinds of Mfg Firms Different ---
36
 MNC’s
 Respond to competitive cost competition by locating where they
need to
 Will locate R&D near production when nexus needed – can go
abroad, avoid risk of adv’d technologies by offshoring to low cost
 Entreprenurial/Startups pursuing scaleable tech’s
 VC’s won’t fund mfg. – not core competency
 So: Offshore production – reduce risk and costs
 But risk product control – knowledge spillover
 Lose next gen technologies
 Small Manufacturers – bulk of US manufacturing
 Need productivity gains to stay competitive
 Lack resources, tech dev/access – so need proven technologies,
processes
 Lack testbeds, financing for productivity options – miss tech waves
37
Remember the Steps:
 #1- Technology advances that yield new
manufacturing paradigms
 #2 – Matrix - Pick new manufacturing technology
paradigms that apply across range of
manufacturing sectors
 #3 – Technology alone is not enough – also need
process and business model
 #4 –Fuse Services and Manufacturing
 #5 – Other nations - Better look over our shoulder
 #6 – Build the Workforce
 #7 – Innovation Organization – Look at those
intermediary stages
 #8 – Build Regional Infrastructure
 #9 – Financing – The 5 vs. 10 year yardstick
Remember: Manufacturing
Scales
38
Economics focused on “decreasing returns”
Brian Arthur – helped us understand “increasing
returns”
 tech advance lock in a standard, geometric
increase
Exs.: railroads, cars, Microsoft operating
system,desktops, iPod-iPad-iPhone,
Services scale slowly – face to face
Manufacturing key to “increasing returns” in an
economy – fundamental to strong growth
Wrap- up:
 Focus here is innovation, the supply side –
the demand side also needs attention:
taxes, trade, regulation -- But: it’s
STRUCTURAL NOT MACRO-ECONOMIC
 Remember the Hourglass - we must
understand the tie between production and
other employment sectors – that’s why mfg.
must be taken seriously
 And Remember, Manufacturing Scales – key
to growth
39