Deploying Computer Aided Design (CAD) systems and measuring their effectiveness is a challenge for any engineering and design team. We first looked at ways to build a “predictive analytics” engine that could measure CAD system efficiency within large groups, recommend training and offer help for individual CAD users and even report CAD crashes and setup issues. This idea quickly evolved into what we call an “AI CAD Consultant”.
As CAD software has continued to evolve from electronic drafting tools in the 80’s to extremely complex engineering tools that offer simulation and generative design capabilities, the effort of all users has evolved as well. From training on simple picks and clicks to the more complex tasks of understanding the true design intent of models, users are inundated with new software.
At the same time, hardware has continued to advance to keep up to the software demand with thousands of computer configurations, graphics drivers and operating system settings to make the entire system run as smooth as possible. This effort alone requires one or more dedicated IT person(s) to work with the engineering and design CAD teams.
With the latest push from software vendors to be subscription based business, features and updates come out constantly. No longer just on an annual basis which has required that user training and IT support is an ongoing process.
Automationforce has over 25+ years experience in all aspects of the CAD world and we looked to use that experience to create a software platform that could assist in the above mentioned challenges. To try and see if we could build a better process than sending all users for the same training, having IT run around helping to sort out issues that make the system crash and essentially ensure that everything runs smoother “automatically”.
We stared out with a central server system following picks and clicks for all users. As the CAD team worked, the system would gather information and use it to look for system irregularities, crashes, user challenges and even file loading times.
While the reporting was helpful, the sheer amount of data was too large to be parsed by a simple database and even as a reporting tool.
To really ensure that we added value to users, we looked at more of a predictive analytics engine. Using the data collected, run machine learning algorithms that essentially get smarter as the more data was accumulated. The machine learning would then predict which users would need training, which system setups would have the best impact and even recommend 3rd party software for the various processes.
We launched our Zoeked platform. An open source CAD Analytics engine that works as your teams own “AI Consultant”.
Side note: The word “Zoek” in dutch means “search” and we add “ed” on the end to mean “have searched” to say Zoeked has searched for answer. Plus we thought it was catchy.
We have been testing the platform with several large CAD installations and the results are extremely impressive. We are in the process of patenting parts of the technology and believe this will soon be available to all CAD teams globally to help them define CAD systems settings, find users the help they need, target micro-training instead of one fits all training and we’ll get to a point where Zoeked will be able to recommend help AS a user is working.