Danesh Tarapore

About

Danesh Tarapore is a robotics researcher whose work spans adaptive robotics, swarm intelligence, and fault detection in multi-robot systems. He is perhaps best known as a co-author of "Robots that can adapt like animals" (2015), a landmark paper with nearly 950 citations that introduced the Intelligent Trial and Error algorithm, enabling damaged robots to recover functionality in minutes by drawing on pre-computed behavioral repertoires — a breakthrough that fundamentally reshaped thinking about robot resilience. Beyond this seminal contribution, Tarapore has built a substantial body of work addressing one of swarm robotics' most pressing challenges: reliably detecting and tolerating faults in decentralized robot collectives. His research spans immunology-inspired fault detection, behavioral outlier analysis in physical robot swarms, and adaptive online diagnosis, reflecting a consistent drive to move swarm systems from controlled laboratory settings into real-world deployments. His 2020 work on sparse swarms directly confronts the gap between academic demonstrations and practical applications. Tarapore has also advanced quality-diversity algorithms, particularly MAP-Elites, exploring how encoding choices and meta-evolution influence the generation of resilient behavioral repertoires. Collectively, his contributions — spanning theoretical foundations and physical hardware validation — have meaningfully accelerated the maturity of autonomous, self-healing robotic systems.

Research Focus

Key Achievements

Top Papers

1. 1
   Robots that can adapt like animals

   948 citations · 2015
2. 2
   Generic, scalable and decentralized fault detection for robot swarms

   54 citations · 2017
3. 3
   Sparse Robot Swarms: Moving Swarms to Real-World Applications

   51 citations · 2020
4. 4
   To err is robotic, to tolerate immunological: fault detection in multirobot systems

   38 citations · 2015
5. 5
   Fault Detection in a Swarm of Physical Robots Based on Behavioral Outlier Detection

   38 citations · 2019
6. 6
   How do Different Encodings Influence the Performance of the MAP-Elites Algorithm?

   36 citations · 2016
7. 7
   Evolvability signatures of generative encodings: Beyond standard performance benchmarks

   23 citations · 2015
8. 8
   Adaptive Online Fault Diagnosis in Autonomous Robot Swarms

   21 citations · 2018
9. 9
   QED: using Quality-Environment-Diversity to evolve resilient robot swarms

   21 citations · 2020
10. 10
    Learning behaviour-performance maps with meta-evolution

    18 citations · 2020

Key Collaborators

Contact & Links