Hardware/IoT
Japan

SCHAFT

$2Klost
6 Years
November 2018
No Market Need
Founded by: Yuto Nakanishi

A world-champion bipedal robotics startup from the University of Tokyo that was acquired by Google, but ultimately shut down after a high-profile sale to SoftBank collapsed due to employee resistance.

The Autopsy

SectionDetails
Startup Profile

Founders: Yuto Nakanishi

Funding: Acquired by Google (later Alphabet) in 2013; previously struggled to find venture capital in Japan

Cause of Death

Market Fit: Commercial Viability: Alphabet struggled to find a commercial path for bipedal robots, which were seen as high-cost research projects rather than products.

Other: Failed Acquisition: Alphabet agreed to sell SCHAFT to SoftBank in 2017, but the deal collapsed in 2018 because several key SCHAFT employees refused to work for SoftBank. Leadership Vacuum: The departure of Andy Rubin (Google's robotics lead) in 2014 left the 'Replicant' division without a clear vision or protector.

The Critical Mistake

Acquisition Dead-End: Tying the company's fate to a single buyer (SoftBank) and failing to address the 'human factor' (employee consent), which left Alphabet with no 'Plan B' other than liquidation when the deal failed.

Key Lessons
  • Talent is the primary asset in robotics; if the engineers refuse to move, the intellectual property often loses its value in an acquisition
  • Bipedal robots are a 'long-game' technology; without immediate industrial or consumer use cases, they are vulnerable to corporate cost-cutting
  • Research-heavy startups must maintain a degree of independence or a clear commercial roadmap to survive being swallowed by a tech giant's shifting priorities

Deep Dive

SCHAFT was the 'quiet overachiever' of the robotics world. While Boston Dynamics gained fame through viral videos of backflipping robots, SCHAFT won the prestigious 2013 DARPA Robotics Challenge (DRC) Trials, outperforming the best labs in the world in disaster-response tasks like climbing ladders and clearing debris. This victory was what caught Google's eye during its massive 2013 robotics shopping spree. The Google 'Replicant' Era Under the leadership of Android creator Andy Rubin, Google acquired eight robotics startups, including SCHAFT and Boston Dynamics, to form a secretive division called 'Replicant.' SCHAFT was moved from Tokyo into Alphabet's 'X' (moonshot factory). For a few years, the startup went dark, operating in total secrecy. They occasionally re-emerged to show off breathtaking tech, such as a low-power, high-efficiency bipedal robot that could traverse rocky terrain and carry 60kg, aimed at civilian and disaster aid. The SoftBank Fallout By 2016, Alphabet's appetite for expensive, non-revenue-generating hardware began to sour. They successfully sold Boston Dynamics to SoftBank in 2017 and intended to include SCHAFT in the same package. However, the deal hit a bizarre and terminal snag: several key SCHAFT employees reportedly refused to join SoftBank. In the specialized world of high-end robotics, the 'human capital' is the product; without the core engineering team, SoftBank walked away from the SCHAFT portion of the deal in 2018. The Final Shutdown Alphabet spent months trying to find a new buyer for SCHAFT, but the market for experimental bipedal robotics was narrow. Faced with mounting losses in its 'Other Bets' category and no suitor in sight, Alphabet decided to 'pull the plug' in November 2018. The company was dissolved, and while Alphabet offered to help employees find other roles within the company, the world-leading bipedal technology was effectively mothballed. The Aftermath SCHAFT's death serves as a warning in the Hardware/IoT and Aerospace/Defense sectors: technological dominance (winning DARPA) is no shield against corporate restructuring. It also highlighted the cultural friction that can occur when a small, academic-focused research team is forced to navigate the high-stakes M&A world of global conglomerates.

Key Lessons

1

Talent is the primary asset in robotics; if the engineers refuse to move, the intellectual property often loses its value in an acquisition

2

Bipedal robots are a 'long-game' technology; without immediate industrial or consumer use cases, they are vulnerable to corporate cost-cutting

3

Research-heavy startups must maintain a degree of independence or a clear commercial roadmap to survive being swallowed by a tech giant's shifting priorities

Share: