FAPESP Conference
”Creating Local Prosperity Through World-Class Science-Based Business Development”
Sao Paulo, Brazil, June 16, 2016
The Startup
Scaleup
Challenge for
Innovation
William B. Bonvillian
Director, MIT Washington Office
What’s the policy challenge and the emerging policy?
• What’s the US innovation system problem?
• Financial system model can’t measure value of modern
firm, i.e., its innovation capacity
• No metrics, so substitute a proxy: short term profitability
• Finance therefore drives US firms to focus on “core
competency,” to reduce vertical assets, go “asset light”
• This drives a reduction in firm research (basic and applied)
• Firm research fell to 26% in 2010-13 from near 30%
from 1985-2000 (as a % of total business R&D - ITIF)
• Firms do incremental development focus instead
• But for breakthrough research for innovation – large firms
increasingly buy startups2
The Startup challenge:
• But this “buy-up startups” option for big firms is
running into trouble:
• There is a decline in finance support for startup model
in most technology areas (other than software/IT)
• Creates a gap in the US innovation system
• Scale up challenge for emerging technologies into
production and implementation
• What policies are emerging? How can we fill this
gap?
• New “Innovation Orchards” concept
• Link startups and small manufacturers (US
Manufacturing Extension Program)3
Emerging Technology: Complicated story
U.S. likes to stand up new
technologies in “New Frontier”
Sectors – in Unoccupied Territory
– Limits competition in the
early technology stages
But most tech is in “Occupied
Territory” – New technologies
have to parachute in and will
be shot at on the way in
-- We do biotech, we don’t fix
health care delivery
4
Related Problem: “Legacy Sector” Issuefor New Technologies
• New technologies that open new sectors typically introduce
“new functionality”
• There was nothing quite like computing before computing
• Can command premium price at entry because of new functionality
• Then over time drive down the cost curve
• But most technologies in complex established Legacy Sectors
• Initial technologies introduce limited new functionality
• Compete against established technologies
• Must compete on price on day one –
• No time to drive down cost curve
• Legacy sectors: well-defended castles
• Protected by technological-economic-political-social paradigms
• Legacy Sector R&D low – ex: Energy industry R&D = > 1% of
Annual Revenue – we will need startups for new energy technologies
•
5
VC’s don’t like Legacy Sectors: Decline in VC Investment in Energy Startups
6
Non-IT Technologies: higher risk/ lower return so no VC• From 2004-2008 VC new energy investment grew from $1B
to $5B (47%/yr.) – Then: After 2008: down to $1B • Energy: high risk, low return vs. software low risk, high return
• 2006-2011: $25B in clean tech VC $; less than half returned • (B. Gaddy, et al, “Venture Capital and Cleantech”)
• MIT Production in the Innovation Economy study - 2014 • Looked at 60 Boston-area startups that survived for more than
decade (focus on non-IT, BT)
• VCs stayed in past normal 5/7 years, didn't want to marginalize holdings, but put firms on income maintenance
• When startup got well into product design, called VC
• The VC reply: “we don’t do this, call our contract manufacturer or sovereign wealth fund in…” (you fill in the blank) • (E, Reynolds, et al, “PIE,” vol. 2, chapt. 4)
7
Enter: Advanced Manufacturing Institutes
• The Issue: U.S. manufacturing in decline –
• hollowing out, employment cut 1/3 from 2000-2010 –
• affects U.S. innovation system because production is creative stage
• The Fix: advanced Manufacturing Innovation Institutes (MIIs)
• Advanced Manufacturing Partnership 2011 and 2014 reports: apply a Fraunhofer model
• 8 Institutes named; 2 through DOE (power electronics, advanced composites)
• But: Mfg. Institute model speaks to existing ecosystem –
• Collaborations of SMEs and major mfg’s. w/univ. and gov’t
• The Startup Scale-up problem isn’t really reached by Institutes
• What to do? 8
Another Answer: SunShot
• DOE’s SunShot program - national collaboration set up in 2007 to make solar competitive with fuels by 2020
• Goal: 6 cents/KW hour total, installed cost
• Have reached 70% of that goal; 30% to go
• Approaches: Driving down production costs, systems integration, driving down soft costs and installation (64% of total costs), incubator and startup programs
• Support for 61 startups since 2007
• $138m in early stage federal funds; leverage $3B in private support
• Includes: hardware innovation and manufacturing, software, cost-reductions throughout the value chain, and capability-related solutions – federal role: many stages, not just R&D
• Now: 2015, disruptive solar technologies (beyond solar PV) 9
SunShot’s Approach:
10
X
X
XX
X – where SunShot helps
Trend in Price Reduction for PV & Capacity Growth, 1980-14
11
Now What? for Startup Scaleup
• But SunShot is solar only and goal is price not helping startups
• And startups are outside the Manufacturing Institute model
• Is there a more systematic approach? Finance?
------
• US AMP 2014 advanced manufacturing study: • recognized the innovation system gap in financing for
scale up for startups • tried to find new financing mechanisms –• not a good time for expensive new programs - couldn't find
a workable one
• Now What?12
“Innovation Orchards”
• Rafael Reif, MIT - substitute space for capital
• “Innovation Orchards”
• Already rich innovation cluster in Boston, but scaleup issue
• Need: technology and equipment rich space for startups
• full of know-how – use regional support to build
• for: advanced prototyping, demonstration, testing - perhaps
small lot pilot production
• Gets startup down scaleup curve
• Derisking – get them into range of more traditional financing
• Could accelerate the innovation – better than VC fix13
Example: Cyclotron Road/LBL
• ModHOMEPROGRAMCOHORTPROJECTS
• MOSAIC MATERIALS
• VISOLIS
• OPUS12
• SPARK THERMIONICS
• CALWAVE
• POLYSPECTRA
• ABOUT US
• MISSION
• TEAM
• GET INVOLVED
• LATEST
• NEWS
• REPORTS
• PRESS
• INSPIRATION
• APPLY
• style below blocks out the mobile nav toggle only when nav is loaded
14
• IDEA: Park a bunch of energy tech
startups (16) outside fence of Lawrence
Berkeley Lab in intermediary space
• Pay them salaries
• Give them the keys – let them invade
• Access to advanced technology,
equipment, know-how
• New tech transition model:
- Old model – inside/out - get the
labs to transition their own technologies
- New Model – outside/in
• Better model? More DOE labs will do
Example: TechBridge• TechBridge – arm of Boston Fraunhofer for energy startups
• TechBridge Role: Prototype, Pilot Production, Test, Validate
• Link startups with industry partners, then prototype. test and
validate technology for production –
• use Fraunhofer facilities for test/validation,
• use industry partner who wants the innovation for scaleup
• So high grade industry level tech validation w/ scaling with
partner – Validation is key
• Match-udep model
15
Does US law allow federal labs to help startups?
• IDEA – federal labs are underutilized assets
• for adv’d equipment. technologies, know how –
• get better utilization from them for startups –efficiency issue
• new outside/in tech transfer approach – low cost
• Can federal labs do this? Yes – Stevenson-Wydler Act of 1980 allows “Partnership Intermediary Agreements” (PIAs)
• state-ok’d non-profits (ex., state univ’s) to set up intermediary org’s. to link labs with tech transfer options
• can include startups who can use facilities
• DOD, NIST, Ag use widely
• SO: PIA model - allows entry to labs and tech transfer
• subject to export controls, other federal requirements16
But what does a federal lab know about manufacturing?
• A federal lab knows technology research and equipment
• But federal labs don’t know about production?
• So: NEW IDEA – a missing feature for innovation orchards:
• Incubator Greentown Labs in Somerville joined Mass.
Manufacturing Extension Program in 1 year pilot program
• Link scaleup ready startups with small manufacturers
• Don't have to go to the prototype shops in Shenzhen – can
find capabilities in your own backyard
17
GreentownLabs/MassMEP
18
•
Here’s what happened: IDEA They need each other - startups need
production, manufacturers don’t do R&D, need innovation access
• 46 startups a Greentown; 33 Startups were interested in linking to SME manufacturers from 7 different locations across eastern Mass.
• 83 Manufacturers were interested in working with and connected to Startups.
• 120 Connections were made between Startups and Manufacturers.
• People intensive effort – online matching won’t work
• 10 Workshops held by Greentown Labs
• Over 100 office hours held at Greentown Labs
• 8 outside startup support networks expressed interest in learning from the Initiative
• 16 Partnerships were formed between Startups and local Manufacturers.
• Basic training for startups on production could go online and scale; but personal relationship-building critical – it’s a marriage, love required
• (sources: Micaelah Morrill, Greentown, Peter Rossi, MassMEP)
Wrap-up: Role of Policy in Startup Scaleup
• So: innovation policy challenge: • Need to scaleup innovative startups for production of new technologies
• But financial support for non-IT startups in sharp decline
• Are new policy models emerging to fix this?• New Manufacturing Institutes and SunShot - impt., but don’t quite fit
• But: “Innovation Orchards” idea fits – substitute space for capital• Rich technology, equipment, know-how space for startups
• Help get them through advanced prototype, demo, testing, pilot production
• Cyclotron Road, TechBridge provide working models
• Another missing link: connect startups to small manufacturers • Can link startup incubators to startups with MEPs -- in every state –
Greentown Labs/MassMEP example
• Can this work? Feds – no new money, can’t create new programs• But these policy fixes are already authorized, have low capital costs, better
utilization of existing assets – labs, MEPs
19