Nesa Innovation Labs increases engineering efficiency by 20%

Overview

Sanjay Patel has been a Control Systems engineer in the automotive vehicle industry for over a decade where he has worked for companies like General Motors and Tata Motors. In his career he has used MATLAB to build different types of automotive systems including transmission automation systems, integrated starter generator (ISG) systems and actuator drive systems. 

In the early stages of Sanjay’s career, MATLAB was the de facto tool for model based design. Sanjay suggests that things may have changed, “Back then no other tool could be used for MBD. However, today, I am not so sure that MATLAB is the right solution for most automotive companies.”

Challenge: Moving away from expensive and limiting engineering tools for model based design in automotive

Like Sanjay, hundreds of engineers in the automotive industry have been turning to Collimator as an alternative to MATLAB. Sanjay did so for the following reasons:

Cost Concerns

Many Tier 1 and Tier 2 suppliers have to sacrifice something to be able to afford MATLAB licenses. Sanjay said, “Cost has always been a challenge, especially when working with smaller companies. You could compare the cost of a complete MathWorks license with the cost of an engineer who has a masters and 2-3 years of experience.” 

As a direct result, many of the original equipment manufacturers (OEMs) Sanjay works with have chosen not to pursue model based design unless it was a strict requirement from the automotive manufacturer. This meant that they cannot benefit from the power of MBD including traceability from requirements to implementation and streamlined validation and verification. 

Computational Limitations

“MATLAB is a standalone desktop application therefore its speed is constrained by the processing power available locally,” Sanjay said. This isn’t a problem for large companies that can afford expensive servers to run their simulations. However it can be a challenge for startups and medium sized companies. As a result Sanjay often has to choose between two bad options “either running simulations on the desktop which would take hours or staying at work until late in the night.”

After COVID hit, neither one of these solutions were viable anymore for Sanjay. He said, “Everyone was told to work from home and the company had to ship more powerful desktops to all their engineering employees. These were costly machines, in the range of $5,000 to $10,000.” 

Limiting User Experience

“With a plethora of features packaged together under one single roof, MATLAB is not user friendly, especially for beginners. It has a steep learning curve… It can take a new engineer several months to become accustomed to using the product.” Sanjay said.

Solution: Adopting Collimator as one of their engineering tools for system analysis and design

Sanjay left Tata Motors to start NESA Innovation Labs. His company specializes in providing model based design (MBD) and ISO26262 consulting services for EV design startups. He said, “not long ago people looking for Simulink alternatives had very few options. There were options like OpenModelica and SciLab, but they weren’t user friendly either.” 

That is, until now. 

Sanjay started using Collimator as his Matlab alternative because it offered him a way to build model based design (MBD) and ISO26262 compliant battery management systems without costing an arm and a leg. Here are the things he said about Collimator: 

Powerful Modeling Capabilities

Collimator offers engineers the same tools that MATLAB and Simulink have, including ready to use plant models and code generation for embedded C. It augments those with a Python scripting environment and high performance computing. After using Collimator for a few days Sanjay said, “I’m pretty confident using [Collimator]...I was able to verify the system very easily, and I already have gained trust that the solver works well.”

In addition, Sanjay was able to find and use open source Python libraries to fast-track his development. He said “I looked up a bunch of battery modeling libraries. The one I ended up using is called Pybamm. It was exactly what I needed.”

User Friendly UI

Collimator’s user-friendly and intuitive GUI allows engineers to focus on the things that matter. That is designing their system. Sanjay said, “It took only 10 minutes for me to get my first simulation ready. It is intuitive, [the way] you can drag and drop blocks.” 

He added, “I showed Collimator to my friend who is a mechanical engineer but had never used MATLAB and Simulink. She was interested in learning system design, but never did because she found it difficult.” He continued “She only had one look at Collimator and felt inspired to learn. It looked very easy and everything she wanted was right there.”

Best-in-Class Collaboration

Collimator offers users seamless version control, similar to Mac’s Time Machine feature. Sanjay said this was “one of Collimator’s best features. Scrolling through previous versions was like watching a movie of how the system came together.” 

Before using Collimator, Sanjay and his engineering collaborators would have to allocate a couple hours every week to meet and decide who will be working on which model so conflicts don’t arise. When faced with a conflict, all work had to stop for the manager to review the changes and determine which branch to use. 

Sanjay said, “Because of the limitations, even when some of us wanted to collaborate, it was not done. I have heard of the GitHub integration with Simulink but to be frank, I have not heard of anyone using it. It’s just too complex.”

Unlimited High Performance Computing

Collimator comes with high performance computing right out of the box. Sanjay said, “Many companies have to manage on-prem servers. Some companies even go as far as owning data centers to use for simulations. Collimator takes all of those costs away.”

He continued, “Collimator’s subscription model is very popular because it reduces the risk of us investing in expensive machines. Also, the IT team was happy because [Collimator] is run on AWS servers with top-notch security as an in-built feature.”

“It took only 10 minutes for me to get my first simulation ready… I’m pretty confident using [Collimator]... I was able to verify the system very easily, and I already have gained trust that the solver works well.”

Results: Increasing automotive software development efficiency while reducing research and development costs

Using Collimator, Sanjay and his collaborators will be able to ensure that time and resources are aligned towards building the product. The estimated savings from this could be up to $45K. That is, $36K from using Python to fast-track development and $9K from streamlined collaboration.

Using Python to fast-track development

Sanjay works with approximately 10 engineers on most projects. “At any given time 2 of those were working on developing plant models.” Using Collimator, they can now use open source Python libraries as a starting point and re-allocate the 2 engineers to work on the controller instead. “The controller is how we get paid,” Sanjay said matter of factly.

\begin{array} {|l|l|}\hline \text{Engineer salary in USD} & $18k \\ \hline \text{Number of engineers reallocated} & 2 \\ \hline \text{Estimated yearly savings in USD} & $36k \\ \hline  \end{array}

Estimated savings from using Python libraries

Streamlining collaboration

Sanjay’s collaborators “meet every week for at least 2 hours to determine who will be working on which model so that someone does not accidentally modify something. Conflict resolution takes another 2 hours a week on top of that. This typically happens at the worst possible time - right when a deadline is close.” Using Collimator they can now reduce the meeting time in half, saving about 20 man hours per week.

\begin{array} {|l|l|}\hline \text{Engineer salary in USD} & $18k \\ \hline \text{Working hours in a year} & 2080 \\ \hline \text{Engineering salary per hour} & $9 \\ \hline  \end{array}

\begin{array} {|l|l|}\hline \text{Total meeting time per week} & \text{4 hours} \\ \hline \text{Reduction in meeting time} & 50\% \\ \hline \text{Savings in meeting time per engineer} & \text{2 hours} \\ \hline \text{Number of engineers} & 10 \\ \hline \text{Savings from unnecessary meetings per week} & \text{20 hours} \\ \hline \text{Weeks in a year} & 52 \\ \hline \text{Savings per year in man hours} & \text{1040 hours} \\ \hline \text{Engineering salary per hour} & $9 \\\hline \text{Yearly savings in USD} & \text{\$9,360} \\ \hline  \end{array}

Estimated savings from streamlining collaboration

“I wanted a tool that had the same capabilities that I got with Mathworks plus more. I needed the  flexibility to scale up the number and complexity of my simulations all without breaking the bank… With Collimator, I did not have to compromise.” And with a smile on his face, Sanjay concludes “I can’t believe it’s Friday, it’s 10pm and I’m excited to use this!” 

Find out more about why engineers are turning to Collimator as their Matlab simulink alternative and sign up for free to start designing your system today.