# DARPA DARPA (Defense Advanced Research Projects Agency) is the research and development agency of the United States Department of Defense, responsible for developing breakthrough technologies for national security. Created in 1958 as ARPA in response to the Soviet Union's Sputnik launch, the agency has funded some of the most transformative technologies of the modern era: [[ARPANET]] (precursor to the [[Internet]]), GPS, stealth aircraft, autonomous vehicles, and voice recognition. DARPA operates with unusual freedom—small teams, high-risk tolerance, and minimal bureaucracy—pursuing "impossible" projects that conventional research avoids. The agency's model is distinctive: program managers with 4-5 year terms fund external researchers (universities, companies, labs) rather than maintaining internal labs. [[J.C.R. Licklider]], who directed DARPA's Information Processing Techniques Office (IPTO) in the 1960s, shaped the vision for personal computing and networked systems. DARPA's willingness to fund speculative, long-term research—with acceptance of failure as part of innovation—has made it a model for "moonshot" thinking. The agency renamed from ARPA to DARPA in 1972 (and briefly back to ARPA 1993-1996) to emphasize its defense mission. ## DARPA Model ``` ┌─────────────────────────────────────────────────────────────┐ │ DARPA INNOVATION MODEL │ ├─────────────────────────────────────────────────────────────┤ │ │ │ DARPA APPROACH TRADITIONAL R&D │ │ ┌─────────────────────┐ ┌─────────────────────┐ │ │ │ • High-risk, high- │ │ • Incremental │ │ │ │ reward projects │ │ improvements │ │ │ │ • 4-5 year program │ │ • Permanent staff │ │ │ │ managers │ │ • Risk-averse │ │ │ │ • External performers│ │ • Internal labs │ │ │ │ • Failure accepted │ │ • Failure penalized │ │ │ │ • Small, agile teams│ │ • Large bureaucracy │ │ │ │ • "Impossible" goals│ │ • Feasible goals │ │ │ └─────────────────────┘ └─────────────────────┘ │ │ │ │ ┌─────────────────────────────────────────────────────┐ │ │ │ DARPA FUNDING FLOW │ │ │ │ │ │ │ │ DARPA ──► Universities │ │ │ │ ──► Defense contractors │ │ │ │ ──► Private companies │ │ │ │ ──► National labs │ │ │ │ │ │ │ │ (DARPA has ~220 employees, minimal internal R&D) │ │ │ └─────────────────────────────────────────────────────┘ │ │ │ └─────────────────────────────────────────────────────────────┘ ``` ## Major DARPA Inventions | Technology | Year | Impact | |------------|------|--------| | **[[ARPANET]]** | 1969 | Became the Internet | | **TCP/IP** | 1974 | Universal network protocol | | **GPS** | 1973 | Global positioning system | | **Stealth technology** | 1970s | Radar-invisible aircraft | | **Drones/UAVs** | 1960s+ | Unmanned aerial vehicles | | **Voice recognition** | 1970s | Siri, Alexa precursors | | **Computer mouse** | 1960s | Via Engelbart funding | | **Hypertext** | 1960s | Via Engelbart, led to WWW | | **RISC processors** | 1980s | Modern CPU architecture | | **Onion routing** | 1990s | Tor network basis | ## Key DARPA Programs | Program | Description | Outcome | |---------|-------------|---------| | **IPTO** | Information Processing Techniques Office | ARPANET, time-sharing, AI | | **DARPA Grand Challenge** | Autonomous vehicle competition | Self-driving car advances | | **Strategic Computing** | AI and parallel computing | 1980s AI winter catalyst | | **HPCS** | High Productivity Computing | Supercomputer development | | **Lifelog** | Digital life recording | Cancelled; concepts in social media | ## DARPA Principles | Principle | Description | |-----------|-------------| | **Ambitious goals** | Pursue "impossible" problems | | **Limited terms** | Program managers rotate (4-5 years) | | **External performers** | Fund outside researchers | | **Technical autonomy** | PMs have significant freedom | | **Tolerance for failure** | Not all projects succeed | | **Bridge to deployment** | Hand off to military/industry | ## Timeline | Year | Event | |------|-------| | 1957 | Sputnik launches, shock to US | | 1958 | ARPA created (February) | | 1962 | Licklider joins IPTO | | 1969 | ARPANET goes live | | 1972 | Renamed to DARPA | | 1973 | GPS development begins | | 1974 | TCP/IP developed | | 1993-96 | Briefly renamed back to ARPA | | 2004 | First Grand Challenge (autonomous vehicles) | | 2007 | Urban Challenge won by CMU | ## Key Figures | Person | Role | |--------|------| | [[J.C.R. Licklider]] | IPTO director, networked computing vision | | [[Bob Taylor]] | IPTO director, funded ARPANET | | [[Larry Roberts]] | ARPANET program manager | | [[Douglas Engelbart]] | Funded researcher, invented mouse/hypertext | | [[Vint Cerf]] | Funded researcher, co-invented TCP/IP | | Ivan Sutherland | IPTO director, computer graphics pioneer | ## DARPA vs Other Agencies | Aspect | DARPA | NSF | NIH | |--------|-------|-----|-----| | **Focus** | Defense tech | Basic science | Health/medicine | | **Risk tolerance** | Very high | Medium | Medium | | **Timeline** | Long-term breakthrough | Various | Various | | **Performers** | External | External | External | | **Budget (approx)** | $3.5B | $8B | $45B | ## DARPA Model Influence | Inspired By DARPA | Description | |-------------------|-------------| | **ARPA-E** | Energy research (2009) | | **IARPA** | Intelligence research | | **BARDA** | Biomedical preparedness | | **HSARPA** | Homeland security | | **"Moonshot" thinking** | Google X, corporate R&D | ## Criticisms and Controversies | Concern | Description | |---------|-------------| | **Military applications** | Technologies used in warfare | | **Surveillance tech** | Total Information Awareness program | | **Dual-use** | Civilian/military blur | | **Funding priorities** | Defense over civilian needs | ## References - Belfiore, Michael. *The Department of Mad Scientists* (2009) - Waldrop, M. Mitchell. *The Dream Machine* (2001) - https://www.darpa.mil/ - https://en.wikipedia.org/wiki/DARPA ## Related - [[ARPANET]] - [[Internet]] - [[J.C.R. Licklider]] - [[Douglas Engelbart]] - [[Vint Cerf]] - [[Bob Taylor]] - [[GPS]] - [[Artificial Intelligence]] - [[Computing History]] - [[Innovation]]