Teradyne’s Aurora Project
Solving The Innovator’s Dilemma
Sailu Challapalli, Michael Chu, Annie Kuo, Emily Liu, Arundhati Singh, Erick Tseng
Roadmap Background
Teradyne Motivation & Background Organization
The Innovator’s Dilemma CMOS Windows NT
The Aurora Project Analysis of Aurora Extending the Aurora Methodology Conclusion
Thesis The Aurora Project solved the
Innovator’s Dilemma However, the Aurora methodology
is not a guaranteed solution to the Innovator’s Dilemma
Sources Key interviews
Alex d’Arbeloff, CEO and Chairman Marc Levine, Product Group Manager Hap Walker, Group Engineering Manager Tom Newman, VP Corporate Relations Gordon Saksena, ICD Engineer
Harvard Business School Video interviews Case studies
Bedford plant trip
Introduction: Teradyne Founded in 1960 by MIT alumni Alex
d’Arbeloff and Nick DeWolf Headquartered in downtown Boston Industry leader for Automatic Test
Equipment (ATE) for semiconductor manufacturing
Annual sales of $1.8 billion 8,000+ employees worldwide Also a market leader in software,
telephone, and PC Board testing
Teradyne’s OrganizationAlex d’Arbeloff
Chairman and CEO
Alex d’ArbeloffChairman and CEO
Owen RobbinsVice Chairman and CFO
Owen RobbinsVice Chairman and CFO
George ChamillardPresident and COO
George ChamillardPresident and COO
James PrestridgeVice Chairman
James PrestridgeVice Chairman
Integra Test(Aurora)
Integra Test(Aurora)
Mixed Signal(ICD)
Mixed Signal(ICD)Memory TestMemory TestLogic Test
(VLSI)
Logic Test(VLSI)
Finance and ControlFinance and Control
Sales and ServiceSales and Service
Alex d’Arbeloff Chairman and
CEO of Teradyne Heterogeneous
Engineer Vision Networking Business acumen History of
experimentation
The “Innovator’s Dilemma”
Established firms follow a technology trajectory in accordance with customer demands
These firms are not mobile enough to respond to emerging disruptive technologies
Technology Trajectories
Per
form
ance
Time or Engineering Effort
Sustaining Technology
Disruptive Technology
Key Components of Innovator’s Dilemma
Innovator’s Dilemma Teradyne
Sustaining Technologies: BipolarUNIX
Disruptive Technologies: CMOSNT
Barriers to Innovation: Industry veteranCustomer-driven
Value Network: High-end ATE
Disruptive Technology Characteristics
Cheaper and smaller Initially worse performance Higher trajectory slope Gains initial foothold in low end markets
Potentially Disruptive Technologies in Semiconductor ATE CMOS Windows NT
ATE Hardware:CMOS As Disruptive Technology
CMOS vs. Bipolar ECL Lower speed Lower accuracy Higher level of integration Less power consumption Simpler design
Industry trend toward CMOS
ATE Software:NT As Disruptive Technology Windows NT vs. UNIX
Traditionally less powerful workstations
Not compatible with customers’ installed UNIX base
Reduced software development time and cost due to OLE/COM
Industry trend towards Windows
The Rise of CMOS and Windows NT
Per
form
ance
Time or Engineering Effort
Bipolar, UNIX
CMOS, Windows NT
Key low end niche market was microcontrollers
Roadmap Background The Aurora Project
Motivations and Goals Implementation Results Solving the Innovator’s Dilemma
Analysis of Aurora Extending the Aurora Methodology Conclusion
Market Environment Highly cyclical Semiconductor
Testing Industry
ATE
indu
stry
in
rece
ssio
n
Reco
very
1986 1990 1993 1995
Reco
rd e
arni
ngs
Aurora Project: Motivations Emergence of new technologies:
CMOS as predominant semiconductor technology Windows NT as operating system
Competitors developing new technologies Credence
“It was clear that our industry was recovering very quickly, and at that point I was looking beyond what we were doing to
see what the holes were”
– Alex d’Arbeloff
Aurora Project Investigate CMOS and Windows NT Produce smaller, cheaper testers Target new customers in low-end
ATE market Microcontrollers
Small market share Trade-off accuracy for cost
Aurora Project: Implementation Challenges
Teradyne as an immobile firm Large Established Customer-driven Public company
Key Leaders Marc Levine appointed
general manager Experienced software
engineering manager in ICD
Total Quality Management (TQM) manager
Hap Walker Background in hardware,
software, and tester design
Aurora Project Created as small division
Separate facilities in Bedford, MA Separate budget
Internal hires and new recruits “Start-up” culture
Quick development Reported directly to board of directors
The New Look and Feel...
J973 VLSI Tester J750 INTEGRA Tester
vs.
J750: A Technical Success
Reduced costs by 25% “Zero footprint
system” Elimination of bulky
mainframe and interconnection cabling
95% parallel test efficiency for up to 32 devices
J750 IG-XL test software uses Windows NT
A Financial Success Total orders valued over
$200M in less than 2 years Fastest product ramp in
Teradyne history
1995
Start
of A
urora
Proj
ect
1998
100 Sys
tem
s sol
d
350 Sys
tem
s
sold 500 S
yste
ms
sold
1999
2000
Solving the Innovator’s Dilemma
Innovator’s Dilemma Aurora Project
Identification of DisruptiveTechnologies:
CMOSNT
Selecting Suitable Markets: Microcontrollers
Dependence on Customers: New market chosen
Constraints Imposed By aLarge Corporation:
Creation ofautonomous division“Start-up” culture
Roadmap Background The Aurora Project Analysis of Aurora
Sources of Internal Resistance
Extending the Aurora Methodology Conclusion
Sources of Internal Resistance
Emigration of Talent Customer Satisfaction Technical Feasibility Previous Instances of Failure
Emigration of Talent Open mobility across divisions
Entrepreneurial Empowers employees
Gambling the firm’s best engineers on a risky venture Failure would waste valuable resources
Managers fear loss of talent to the Project Jeopardizing the firm’s core business
Customer Satisfaction: Lack of Customer Interest Testing business is very customer-driven
Based on long, well-developed relationships Teradyne must be very responsive to
customer needs Very little customer interest in Aurora
Most customers required greater accuracy Customers preferred familiar technology
Avoid wasting resources retraining technicians
Customer Satisfaction: “The Customer is Always Right”
Teradyne managers felt obliged to listen to customer needs and recommendations
Managers did not want to devote resources to undesired technology
“These projects are big bets. So you bet the division every time you do one of these
things… The problem is that your customers lock you in.”
– Ed Rogas, Vice President of Logic Test Division
Technical Feasibility Doubts about meeting device
specifications Device Speed Device Accuracy
Software Compatibility Backward compatibility with previous
product line Compatibility with other Integra test
products
Past Failures: History of Experimentation
Alex d’Arbeloff has a history of suggesting new technologies to the firm Kinetrix J401 Tester
Many past projects were unsuccessful Many of these projects were also met
with internal resistance but then initiated with d’Arbeloff’s backing
Roadmap Background The Aurora Project Analysis of Aurora Extending the Aurora Methodology
Past Failures
Conclusion
Aurora’s Extensibility
Is the Aurora methodology a formula for solving the Innovator’s
Dilemma…
1) Within Teradyne?2) In any other company?
Past Failures: Kinetrix Foray into Semiconductor Handling
Business Populated by 12 companies with no
economies of scale to step forward Technology was not developed with
customers in mind Collaborated with MIT mechanical engineers
Customers did not find the device familiar
Past Failures: J401 Tester Project Goals
85% cheaper than existing testers Easier to program
Implementation D’Arbeloff set up a new facility three blocks
away from Boston headquarters Tom Newman hired internally to lead
initiative Startup culture
Past Failures: J401 Tester Results
Successfully met product specifications Product was 75% cheaper Programming interface was extremely
user-friendly No market adoption
Still too expensive for new intended market
No “correlation in test” Sold only 24 testers in first two years
(compare to Aurora’s sales of 350)
J401 Tester: A Direct Correlation to Aurora Both projects were…
Extremely ambitious Met with similar internal resistance Initiated thanks to Alex d’Arbeloff’s
persistence and backing Set up as startup ventures Technically successful
Roadmap Background The Aurora Project Analysis of Aurora Extending the Aurora Methodology Conclusion
Conclusion The Aurora Project did solve
Christensen’s Innovator’s Dilemma Overcame hurdles that are typical of large,
established firms Established a new product trajectory Technology was successfully reintegrated
back into the firm
Conclusion (continued) Teradyne has not developed a
formula for success Too many uncontrollable forces
Market conditions Technical feasibility Company culture
The Alex d’Arbeloff Factor
“It may be that you need a CEO like Alex d’Arbeloff as part of the formula [for solving the Innovator’s Dilemma]. He’s a
very bright, very persistent person and he understood entrepreneurial startups.”
– Marc Levine, Aurora Project Manger
Closing Remarks “Discovering markets for emerging technologies
inherently involves failure, and most individual decision makers find it very difficult to risk
backing a project that might fail because the market is not there.”
– Clayton Christensen, The Innovator’s Dilemma
“One must always be prepared for failure. Failure does not result in professional death. Instead,
we must wrap failures back into the firm.”
– Tom Newman, J401 Project Manager
Questions
Teradyne’s Aurora Project:Solving the Innovator’s
Dilemma