“Collimator is the first tool I pull up when evaluating new projects, especially when I need to collaborate with different teams to get go/no-go decisions. I’m excited that I can finally spend... less time on things that do not matter.”
Grace Connors, a Mechanical Engineer and Graduate Research Assistant at MIT, is passionate about clean water. Her masters research focused on groundwater desalination as a means to provide fresh water for the earth's growing population. Her goal was to design a system that optimized the desalination processes in order to produce the most amount of freshwater for the least amount of energy consumed.
Energy consumption of a water desalination system can depend on several different factors including the battery size, power system used, process used, etc. The model that Grace built used model predictive control (MPC) to determine how much energy was being consumed per unit time and solve for the most optimal parameters that minimized energy consumption. To model and simulate the desalination plant, Grace used MATLAB, Simulink and a handful of other tools.
Grace recently partnered with Collimator to solve the challenges she faced as a researcher. Collimator is a modeling and simulation platform for engineers to design and test complex, mission critical systems in a way that is reliable, secure, fast and intuitive.
Grace ran her models at the Lincoln Laboratory Supercomputing Center at MIT which provides researchers with interactive, on-demand parallel computing. Her workflow entailed creating the models in Simulink on her local desktop then submitting the jobs to the Center for processing. Each new iteration of the model would require “changing at least 7 variables over a very wide range of values and simulating the system over the course of at least a year.” She added that “this meant that I had to run the equivalent of 1000s of simulation years and with high fidelity. I definitely couldn’t do this on my laptop!”
With Collimator, she is now able to access high performance computing (HPC) in the cloud with just the click of a button. Though she no longer has access to the MIT supercomputing services, she can still run equivalent models, getting results in minutes instead of hours. She no longer has to compromise the fidelity of her models and it won’t cost her significantly more money or time to provision additional servers.
Grace’s team was forward thinking and always wanted to use the latest release to build their models. However, each engineer ended up having to spend at least half a day every quarter trying to make things work. “It was super disruptive. At one point I had three versions of MATLAB installed and I did not want to download and share the model in the new version because it would be incompatible with the old version. It was very problematic.” She added “Of everyone I worked with I can only think of two people who successfully installed everything correctly.”
Collimator is versionless. Users can access all toolboxes and modules and don’t have to worry about the next release breaking their model. “The concept of forward and backward compatibility isn’t even relevant any more,” Grace said. She is also able to share her models and results directly within the app so she doesn't have to download a model just to share it with her team or wonder what version her teammate has. Collimator just works.
“Pricing is a big deal for a lot of startups. I was fortunate that as a researcher I had access to the MATLAB and Simulink student licenses but I have not always been so fortunate. I worked at a startup before where we had to choose between either buying PLC software or MATLAB and we certainly couldn’t afford both.” To win a large deal, they were forced to buy the PLC software which meant that they could not simulate the performance of a critical motor.
They instead had to test the hardware itself which required additional engineering time and testing harnesses making it significantly more expensive in the long run.
“Companies in similar situations should not have to compromise mission critical aspects of their product development process.” Grace said. With Collimator, pricing is affordable and easy to understand. In fact, getting started is as easy as putting your credit card information and getting started in minutes.
“The problems we faced were often the difference between success and failure and a solution did not exist then. When I saw Collimator, I saw tremendous potential,” Grace said. She joined Collimator for a few months as an advisor and then went back to hardware design after the product launch.
Today, Collimator is incredibly easy to use and approachable, especially for a first time user. Its signal processing, classical controls suite, and the ability to leverage Python for scripting is unparalleled."
She added “Collimator is the first tool I pull up when evaluating new projects, especially when I need to collaborate with different teams to get go/no-go decisions. I’m excited that I can finally spend more time trying to solve the global water crisis and less time on things that do not matter.”