In the engineering and manufacturing sector, the number of mechanical parts and components are immeasurable. With more and more parts being designed for new products, the library of parts that engineers need to look through is endless. Often, the process involves studying a design that is given in a drawing or PDF format and then estimating the cost of manufacturing, which usually takes a couple of hours at a minimum. Indiana-based VizSeek offers its visual search engine as a SaaS platform, which enables users to run designs or part images through it to find the exact match in the manufacturer’s database.This drastically reduces the amount of time required while retaining search accuracy.
Normally, once a company completes a design, and it serves its purpose, it is relegated to the parts database without ever being re-used. VizSeek uses its visual search engine to help enterprises unleash the full potential of their database and digital assets, while also going to great lengths to ensure that designs are not published, maintaining their security. The search engine uses 3D model, 2D drawing, hand-drawn images, photographs, PDF files, amongst others and matches it against the pre-defined parts database which the manufacturing company has compiled over the years. As a result, the search engine makes the parts searchable, without compromising the security requirements of enterprises. VizSeek can also be configured to return search results with detailed part attribute information, allowing users to order the required parts directly.
The search engine makes the parts searchable, without sacrificing the security requirements of companies
In the large-scale manufacturing industries, like automobile or aircraft manufacturing, around 30 percent of parts account for redundancies. “This either comes from new engineers not knowing the history of parts or companies merge with different divisions and each of them has their own parts design for repeated parts,” says Jamie Tan, CEO of VizSeek. When a company integrates VizSeek in its database, the latter indexes all the models and designs and stores the inventory on its servers. In the process, VizSeek gives the company a list of redundancies as well, which puts an end to the manufacturing of unnecessary parts, further improving their ROI.
Visual search is already in place in the e-commerce industry, in stores like Home Depot and eBay (online), but these engines rely only on the image to get the required information. These engines are often trained on what to look for, with millions of images. However, for industrial application, the lack of labeled images hampers visual search engine training. VizSeek can bypass this, by utilizing other digital assets such as 3D models, 2D drawings and PDF documents .What sets VizSeek apart from the rest is the accuracy, with which it delivers search results, it uses shape and other features to find the exact match in the database.
The accuracy of the visual search was chief among the reasons why a heavy equipment company adopted the solution. Located in a remote area, the company was up against competition that could manufacture spare parts, which ultimately put a dent in its revenue. In a move to improve competitivity, the company began working with VizSeek to reduce downtimes when machinery needed service. The company used VizSeek’s application to quickly identify parts through images and was able to order the replacements immediately, for faster turnaround time.
In the future, VizSeek’s core competency will continue to be the manufacturing and engineering industry, but the firm hopes to build a strong marketing and sales team to begin engaging with companies from different industries. The firm is also looking to increase its team, to appeal to manufacturers from other countries as the sales cycles and support structures differ between countries.