QuITO: Team

Team of developers

  • The first version of QuITO was developed by Nakul Randad, Siddhartha Ganguly, Mukesh Raj, and Rihan Aaron D’Silva.

  • QuITO v.2 was developed by Siddhartha Ganguly, Rihan Aaron D’Silva.

  • QuITO v.3 is being developed by Siddhartha Ganguly, Srivardhan Nandkumar Kondekar, and Gaurav Vallabhdas Revankar; we intend to release it as a Julia package.

  • The development of QuITO benefitted from early discussions with Ravi Banavar and subsequent encouragement from Ashish Hota, Mayank Baranwal, and Rohit Gupta.

  • The later versions of QuITO v.2 benefitted from several scientific comments of the anonymous reviewers of GDC2024.

  • I play a supervisory role.

Special mentions
  • An anonymous reviewer of GRDRC2023, through their singularly negative comments, spurred a comparative study of QuITO against some other available techniques. The more disparaging their comments became over successive reviews, our studies unearthed an ever-growing number of quantitative benefits (several orders of magnitude in cerain cases!) of QuITO over other techniques; we reported some of these findings in both GRDRC2023 and GDC2024 in the interest of science. We thank the reviewer for their comments.

  • An anonymous reviewer of GDC2024 raised the issue of the apparent inadequacy of the first order nature of the error in QuITO v.2, and wrote that its performance would be “... highly inferior compared to direct collocation or DMS methods with higher convergence order”. Perhaps they missed the fact that two different metrics were being compared in their statement since their “higher convergence order” is measured relative to the \(L_2\)-metric whereas QuITO v.2 works with the uniform metric. GDC2024 contains several illustrations that demonstrate that QuITO v.2, with its modest first order uniform error control, significantly outperforms the indicated higher order methods in several benchmark singular problems; we invite the readers to check the validity of the italicized phrase in the preceding quotation in other optimal control problems. We thank the reviewer for their comments.

It is somewhat unfortunate that most of our reviewers appear to miss a key point of departure of QuITO from conventional techniques — that of employing the uniform metric from the ground up.