Would you prefer a large company or a great company?

The automation project proposed for implementation in Data Center operations requires a reliable and robust technology that enables 24/7 operation in these critical infrastructures, such as autonomous driving applications, for example.

A person needing to use an ATM likely cannot accept a message on the screen saying that assistance will be available in 45 minutes due to the application being offline.

Ultimately, reliability is the foundation for the application of automation techniques to this type of infrastructure.

In any automation process, monitoring elements are required. These elements essentially indicate the state of something and display a result that shows whether the action taken is as expected or not.

This equipment can sometimes become a hindrance to automation due to the high investment costs involved or because industrializing the solution may be unfeasible.

In this research and innovation project, we aim for automation to present real feasibility. To achieve this, reducing economic requirements is a key component of our proposal..

Bjumper specializes in best operational practices in Data Centers. That's why we implement technology that can perform this work more effectively and efficiently..

With this approach, we consider that automating the climate control process in the Data Center, both in new and existing ones, will require a high investment in monitoring equipment. Therefore, we propose making a simple inference between the electrical consumption of the IT equipment, and thus the heat they generate, with the temperature that can be reached in the rack due to this reason..

 This requires electrical monitoring of each rack, which is quite common in newly constructed Data Centers. However, it's something that many existing centers lack without making a significant investment and assuming scheduled downtime for the installation of measuring equipment..

Going deeper into this relationship, we consider the inverse objective. If we have that relationship between consumption in KW, which converts into heat and therefore into a certain temperature, why can't we perform the reverse calculation??

Let's say we have a rack with IT equipment consuming an undetermined amount of energy. Why not estimate this consumption in kW based on a series of temperature measurements at the front and rear of the rack?.

 The good news is that it's possible. By comparing the real measurements taken from the toroids of the breakers supplying the rack and making measurements and corrections to the mathematical estimation, we can approximate that in a well-structured room—meaning with properly organized rows, relatively symmetrical aisles, racks of similar heights, using blank panels in empty rack spaces, etc.—the approximation model is quite reliable. It allows for estimation, which in many cases is more than sufficient to manage reasonable operations in existing infrastructures without a significant CAPEX requirement..

Therefore, we can start thinking that not only the latest large-scale data centers from the giants can be more sustainable, but also older and smaller infrastructures can move in the same direction with reduced investment and in record tim

This is the essence of Bjumper: turning knowledge into useful, accessible, and affordable tools to revolutionize Data Center operations from a humble company.

With this vision, the working team composed of Bjumper, Tychetools, Trevenque, and the University of Castilla la Mancha takes on the challenge of automation in Data Centers in this DC AIOps project..



A Practical Guide for Incident Management in Data Centers