Blog post

Unveiling the CTO’s 25-year journey: from startup passion to code innovation in simulation

Written by Luka Onesti

28 November 2024 · 10 min read

Looking back on my 25-year journey at ESTECO, one thing stands out: at heart, I’m a startup guy. Pushing the boundaries of what’s possible in software engineering has always been the driving force behind my career. It’s no surprise that this has also been powering our company's growth since 1999. My own journey, however, started even earlier.

Over the years, I’ve had the privilege of contributing to transformative initiatives, with a particular focus on bringing code innovation in engineering design optimization software and beyond. Let me share some pivotal moments and challenges - from the early 90s to recent advancements in simulation technology that empower us to realize our vision at ESTECO: to make multidisciplinary simulation, optimization, decision-making and business process management easily accessible to everyone, to model any enterprise, social or human activity.

Pioneering the use of Java language in the middle of 90s

I graduated in electrical engineering from the University of Trieste (Italy) in 1995, the same year the Java programming language was released. Write once, run anywhere (WORA) was the slogan to illustrate the cross-platform benefits of the Java language. Ideally, this meant that a Java program could be developed on any device, compiled into standard bytecode, and be expected to run on any device equipped with a Java virtual machine. This was especially groundbreaking for web application development. At the time, Fortran was a dominant programming language in scientific and engineering communities, but it was tied to a single machine. My thesis project, however, was written in FORTRAN 77. I moved on to the Parallab department at the University of Bergen (Norway) on a scholarship from the Italian Ministry for Foreign Affairs. There, I worked on a research project that focused on the parallelization of a local area ocean circulation model using High Performance Fortran built upon FORTRAN 90.

A year into my research fellowship in HPC software and system technology at Parallab, I had the opportunity to collaborate once again with my thesis supervisor, Enrico Nobile, who was a young associate professor at the University of Trieste, and Carlo Poloni, a researcher at the same university. Together, we started working on an European funded project called Frontier - Open System for Collaborative Design Optimization Using Pareto Frontiers, led by British Aerospace. The project spanned four years and at the end of the exploitation phase we founded ESTECO as a spin-off of the University of Trieste. This marked the first time we used Java to adapt code to develop an engineering design optimization software. Specifically, one of the first tasks was to migrate the optimization algorithms, which were originally written in FORTRAN by Carlo Poloni, to Java. This transition involved porting the existing codebase to make it compatible with web-based applications and improve its accessibility. Fun fact: one of the project's expenses was purchasing all five Java manuals available at that time.

In addition to the Frontier project, I was also part of the founding team of ICE, a spin-off of the Parallab department of the University of Bergen. Once again, we used Java to build components like the ICEBrowser, which was one of the first ever Java-based web browsers. The spin-off was acquired by the American company Windriver and ICEBrowser was installed in hundreds of thousands mobile phones. That was the late 90s. Today, the ICEbrowser remains part of the Oracle ecosystem.

Testing Web technologies at ESTECO since 1999

1. FRONTIER 1.0 coded to run on Netscape browser

The product was named modeFRONTIER in 2003 when it was first distributed in Japan. Prior to that, it was simply FRONTIER. In 1999 the software was still very experimental and lacked a graphical workflow. In fact, FRONTIER 1.0 was ahead of time: it was developed in Java 1.0 and 1.1 and designed to run on the Netscape browser. Our goal was to bring innovation to the computer-aided engineering (CAE) world, where most software was still written in FORTRAN. We aimed to anticipate what we and other software providers would achieve two decades later with web-based platforms for simulation-driven product development. A few early adopters had the opportunity to test its web browser and run design evaluations, including some of our industrial partners of the Frontier project such as Daimler-Benz Aerospace, Defence Evaluation and Research Agency, Electrolux-Zanussi and British Aerospace.

2. SP4web

The limited functionalities offered by web 1.0 changed our approach. In the early 2000s, modeFRONTIER was completely rewritten in Java 2.0 and 3.0 to run as a desktop application. However, our passion for web technologies remained strong and we continued to explore different ways to bring optimization and simulation to the web. One of our projects was SP4web which started in 2003. It was designed to offer a selected audience access to a range of computational services through a browser interface. Specifically, it enabled users to easily create and deploy complex computational and optimization services - services which typically required the interaction of several computational tools, and, at times, human intervention.

3. Sail4web

A similar venture in this field was Sail4web (2005). In a previous post about fostering innovation in sailing with design optimization, Carlo Poloni explained how Sail4web aimed to provide yacht designers and sail makers with a web application that allowed them to use numerical methods for fluid field prediction in an easy and reliable way. The tool also featured a database browser, a project editor and a post-process environment.

Expanding ESTECO's digital engineering solutions with AI, cloud, BPM, SaaS, and mobile apps technology trends

1. modeFRONTIER has evolved with AI-powered algorithms

Around 2000, we introduced the first graphical user interface in modeFRONTIER 2.0, built using Swing, a widget toolkit for Java. This interface allowed users to create graphical workflows, integrate different simulation tools and apply our proprietary Multi-Objective Genetic Algorithm (MOGA-II) to explore the design space. With optimization algorithms driving the design process and taking the boredom of the trial-and-error approach out of the picture, at the time modeFRONTIER could be considered an Artificial Intelligence precursor. Over the past 25 years, the software has undergone a remarkable evolution driven by a robust code base and continuous development to solve computational heavy multidisciplinary and multiobjective optimization problems. AI has remained a core focus throughout this journey. We’ve prioritized providing users with a simplified optimization experience, offering fast access to set-up simulation workflows and configure optimization algorithms. And, enriching modeFRONTIER with a vast array of AI-powered algorithms to train CAE models and make fast design predictions. By the time AI became a buzzword, we were already ahead of the curve.

2. SOMO/VOLTA platform: from web to cloud deployment

Web 2.0, which was built around the idea of the web as a platform, along with the trend toward shared production platforms and modular systems in the automobile industry, contributed to the launch of our web-based Service Oriented Multidisciplinary Orchestration (SOMO) platform in 2013. The timing was perfect and our original vision of running simulations on the web within a collaborative engineering framework finally hit the market.

By working closely with Ford, one of our main customers in the US, we gained an in-depth understanding of the governance of engineering procedures and the need for collaborative design optimization technologies. SOMO was the first product of this experience, designed to manage the complexity of running multidisciplinary design optimization (MDO) projects with its collaborative framework for multi-domain simulation workflow management.

Since then, our web-based platform has evolved into its second generation: VOLTA, the digital engineering platform for simulation process and data management (SPDM) and MDO. With VOLTA, we aim to shape the present and future of simulation-driven design product development, guided by three core principles: user experience (read more about SOUL, our design system), cloud service architecture and hyperautomation with business process management (BPM).

3. BPMN 2.0 standard for modeling engineering business processes

Standardization is another key concept we’ve been pursuing for years. Nowadays it’s essential to deliver high-quality and scalable technology, but we first stumbled upon it more than 20 years ago. In 2002, I was attending a business meeting on design optimization at General Motors in Detroit. There was an ongoing discussion about adopting formal process models based on semantics like XPDL (XML Process Definition Language), SysML (systems modeling language) and UML (unified modeling language) for any simulation workflow representation. None of these were suitable for the task at hand. SysML offers graphical representation only, while UML provides developers with a standard way to visualize and build new systems. So, together with some of my colleagues, I started to look at different standards like business process modeling language (BPML), the precursor of business process model and notation (BPMN) and the business process execution language for web services (BPEL4WS). The latter was overly complicated and only had XML format representation.

In 2005, the object management group (OMG) introduced the BPMN standard. This quickly became the de-facto standard for business process modeling since it couples a graphical representation with XML machine readable semantic. In 2013, our R&D department developed BeeBPMN, a web editor based on the BPMN standard which contributed to the EU funded Composelector project. This project aimed to develop a business decision support system (BDSS) which integrates materials modeling, business tools and databases into a single workflow. Later, our in-house BPMN editor was rebranded as Cardanit - the ESTECO BPM SaaS solution - reflecting the addition of the Decision Model Notation (DMN) standard. Building on the know-how gained from developing Cardanit, we leveraged our deep understanding of BPM technology to integrate it seamlessly into our VOLTA digital engineering platform. This integration empowers our users to manage the entire engineering process, orchestrate computational tasks and human activities effectively, and optimize the design process.

4. Enterprise mobile apps

Gartner predicted that the demand for mobile app development services would grow at least five times faster than internal IT organizations’ capacity to deliver them, by the end of 2017. At that time, I was worried about the potential dominance of mobile apps over the traditional web browser. This uncertainty prompted us to investigate how mobile apps could support our digital engineering technology. Diwiner, now called Vinnie, was our first mobile app prototype designed to assist with wine and food pairing. It combines user generated contents, such as wine properties and subjective tasting experiences, with a dedicated machine learning algorithm to help users select wines or match food with wines. The concept behind Vinnie can be applied to many other fields where a database is available, including our traditional engineering domain: simulation and experimental data. Over time, the knowledge we acquired from mobile app development became a foundational pillar for the creation of our digital business platform. This platform will introduce all ESTECO technologies in a formal SaaS model, revolutionizing how we showcase our products.

5. Our next digital business platform

Our passion for Artificial Intelligence, standardization, web and mobile technology has been there from day one, and brought us to where we are today. But in the era of digital business platforms, the time has come to bring all these pieces together, adding one more concept we've always aimed for: appification.

We recognize that working within a common digital business platform offers numerous advantages that can enhance organizational operations. By housing applications and data in one location, companies can streamline operations, leading to improved efficiency and cost savings. Users can benefit from standardized concepts across all aspects of the platform, which can lead to increased productivity and collaboration within the organization. And, a shared business data infrastructure underlying all tools and applications eliminates the need for time-consuming integration.

As part of our strategy to broaden the applicability of ESTECO software solutions, we started developing our digital business platform that will be able to embrace our existing (modeFRONTIER, VOLTA and Cardanit) and future software solutions. Within the ESTECO digital business platform every app will be deployed once and be reused across multiple products, making it viable within a SaaS framework. This approach is key to delivering tailored solutions for different user needs while maintaining consistency and maximizing the quality of the user experience.

Our users at the center of everything we do

For over two decades, since our founding in 1999, we’ve been at the forefront of innovation in the field of simulation. Our aspiration is to continue to innovate well into the future - at least until 9999. Our creative spirit and commitment to shaping the future of simulation technology together with our users is at the heart of everything we do.

As CTO, I encourage our software developers to harness the transformative power of AI. From deploying AI/ML models in modeFRONTIER, VOLTA and Cardanit, to using generative AI for ESTECO software design and development and AI-powered virtual assistants, we’re accelerating our journey toward achieving our mission; our goal is to deliver a comprehensive private and public cloud technology stack for multi-disciplinary design, simulation, optimization, and decision-making activities, to effectively manage both our existing product portfolio and all future ESTECO applications.

Luka Onesti
Luka Onesti

Luka Onesti is the co-founder of ESTECO and the company's Chief Technology Officer. He graduated in Electronics Engineering at the University of Trieste and has an MBA from MIB School of Management. He collaborated as a visiting researcher at the University of Bergen (Norway), where he took part in the “Frontier” project. He later co-founded ESTECO and participated in the original modeFRONTIER development.

Luka Onesti is the co-founder of ESTECO and the company's Chief Technology Officer. He graduated in Electronics Engineering at the University of Trieste and has an MBA from MIB School of Management. He collaborated as a visiting researcher at the University of Bergen (Norway), where he took part in the “Frontier” project. He later co-founded ESTECO and participated in the original modeFRONTIER development.

Celebrate 25 years of looking to the future

Discover more about 25 years of experience enabling innovators to transform their products and processes.

Celebrate 25 years of looking to the future

Discover more about 25 years of experience enabling innovators to transform their products and processes.

Learn more
Celebrate 25 years of looking to the future

Discover more about 25 years of experience enabling innovators to transform their products and processes.

Learn more