How We Built a Pop-Up Factory in Just 40 Hours at The Digital Factory 2023

In today’s rapidly evolving and increasingly digitized landscape, manufacturers of all sizes are accelerating their digital transformation efforts in order to become more agile and resilient.

The past few years of tumult in the industry have made this imperative clear as day. In order to stay competitive in this new environment, businesses need to be able to adapt quickly. This requires them to architect their tech stack around an open ecosystem of technologies without limits on the kinds of solutions they can build. But how exactly are manufacturers supposed to create a cohesive solution to their production woes when there are so many disparate pieces of software and hardware available? And how can they implement these technologies quickly and drive real change in their operations rather than simply spinning their wheels in pilot purgatory? That’s where Tulip’s Frontline Operations Platform comes into play.

We teamed up with Formlabs and Autodesk to show just how quickly and easily Tulip can be implemented into any process. Our Pop-Up Factory allowed attendees at The Digital Factory conference to step into the shoes of a frontline operator and experience the same digital tools and processes that form the basis of digital transformation in manufacturing. Using a Tulip app to drive the workflow, attendees were guided from start to finish through the process of building their very own wireless mouse, which they then got to bring home at the end of the day. Read on to learn how we leveraged Tulip to seamlessly integrate a wide range of different technologies and streamline the production process for attendees at our Pop-Up Factory.

With limited time to plan, teams across Tulip, Formlabs, and Autodesk quickly came together to choose the item to be manufactured. The design requirements were that the item would need to be easy for operators to put together, showcase advanced manufacturing techniques, and allow for production to be quickly scaled up should the need arise. Having chosen a wireless charger for last year’s Pop-Up Factory, we decided to continue the theme of wireless personal electronics.

Ultimately, we settled on a bluetooth mouse made with snap-fit, 3D-printed parts. This decision came down to the straightforward assembly process, availability of the necessary electronic components, and opportunity to create a visually appealing yet highly functional item. Plus, we thought attendees would appreciate being able to take a closer look at the internal engineering of an object with which they interact on a daily basis.

However, choosing what sort of item attendees would take home was just the first step. Next, we needed to design the production process that attendees would follow to assemble the mouse during the event. We had to sit down and solve a number of production challenges, such as:

• How will we actually design the 3D-printed enclosure to be economical but also easy to assemble?

• How can we print enough parts at scale in order to have enough for the event?

• How will attendees be guided through the assembly process?

• How can we ensure that attendees with zero hands-on manufacturing experience assemble their mouse without making errors?

• And how can we pull it off with just days to go?

Using Autodesk Fusion 360, the team at Autodesk worked to develop our mouse from the conceptual stages all the way through design verification and manufacturing with their end-to-end CAD package.

Especially pertinent for our use case was the extensive suite of tools available to aid in designing for additive manufacturing (DfAM), as well as for generative design. These included the ability to create lattice structures to increase stiffness and lightness of printed parts, automatically arranging multiple parts to fit within a single build volume — as well as simulation tools to analyze how the design would perform under a variety of test conditions.

We further leveraged Autodesk’s suite of tools to add an element of individual customizability to the mice by adding a relief pattern to the surface of each mouse. These ranged from simple voronoi noise, to a volumetric lattice, and even zebra stripes.

Having finalized the design of our mouse components in Fusion 360, we were able to move on to the mass production of the parts that attendees would later be assembling.

In order to highlight the extreme flexibility offered by 3D printing, we decided to offer attendees the choice of three materials with which they could construct their mouse. The specifics of each material determined how we would go about producing the parts:

• Nylon-12, Unfinished SLS: Similar to more traditional methods of additive manufacturing, Selective Laser Sintering (SLS) allows for the creation of parts with complex internal geometries impossible to achieve with traditional subtractive processes. The SLS technology employed by Formlabs also enables the creation of finished parts without the need for supports or long curing times required by traditional FDM or SLA printers. The particular material we decided to print in was Nylon-12, selected for its high tensile strength, ductility, and environmental stability.

• Nylon-12, Tumbled SLS: While the hard-wearing surface finish of a SLS part straight out of the printer is well suited for prototypes, manufacturers often implement additional steps in order to improve its look and feel for products that will actually end up in customers’ hands. Attendees had the option of choosing an SLS print that had undergone vibratory tumbling and then been dyed in order to provide a more consumer-friendly appearance. Despite starting exactly the same as the unfinished Nylon-12 parts, low-cost post-processing like tumbling can reduce surface roughness by up to 80%.

• BioMed Clear, Medical-Grade SLA: Stereolithography (SLA) printing is one of the most common methods of additive manufacturing, and is typically used when high print fidelity is a priority. To showcase this technology, we also provided attendees with parts printed in BioMed Clear, a transparent, FDA-approved resin typically used in biomedical device applications. The resins used are specially formulated, and final parts are typically sterilized in order for them to be safely used in everything from dental aligners to joint replacements.

With hundreds of attendees set to make their way through the Pop-Up Factory production line, we needed to significantly scale up production in order to meet the expected demand at the event. To accomplish this, we turned to Formlabs, who leveraged their internal print farm in order to produce the components we would need. While printing hundreds of these parts would typically be a very tedious process, thanks to Formlabs’ automation tools we were able to print all the parts we needed in time. With new tools like the Form Auto to enable 24/7 printing, they were able to quickly print the huge volume of parts required with minimal human intervention necessary. Integrating the Formlabs ecosystem into your workflow with Tulip allows you to take this one step further by connecting to their API in order to send files, monitor consumables, and track prints from start to finish across a fleet of printers.

Of course, we still needed a physical workspace where attendees would actually be doing the work of assembling their mice. Our set of requirements were straightforward: Each station would need an area to build the device, bins to hold our inventory of parts, as well as mounting points for a monitor, machine vision cameras, and other IIoT devices used in the production process.

Our completed design for the production line consisted of four identical assembly benches, each with a mounted screen to display work instructions, a Tulip light kit to guide picking, two off-the-shelf cameras for computer vision, an andon light to show station status, a foot pedal for operators to interact with the app, and a Tulip Edge IO device to provide connectivity.

Having fully assembled the workstations, our wireless mouse production line was almost ready to swing into full gear.

All that remained was to connect the various hardware and software elements together with an intuitive user interface to guide attendees through the production process. To accomplish this, we leveraged Tulip’s Frontline Operations Platform (as well as the Edge IO), which enabled us to seamlessly integrate everything from Andon lights, label printers, and machine vision cameras into our app in order to bring our production line to life.

At its core, Tulip is a no-code platform — empowering citizen developers to build powerful solutions like these without needing to tap limited IT resources to program a custom app from scratch. This significantly accelerates the development cycle, meaning that projects implemented in this way can achieve a much shorter time-to-value. Case in point: The apps we built to run our Pop-Up Factory were developed in just 40 hours!

Tulip further simplifies the process of integrating into your existing infrastructure thanks to its connector framework, which enables straightforward integration with any open API. This means that whether you’re looking to connect to your legacy machines or implement the latest generative AI tools, you can do so in Tulip without needing to expend valuable engineering resources building a custom solution from scratch.

On the morning of the conference, attendees began streaming into the convention center, and we now had the labor we needed to run our production process. However, unlike your typical plant operator who may have years of experience manning their particular part of the production line, our workers would effectively be going in blind. In order to ensure that the mice were still being assembled without any defects, we created an app in Tulip to guide attendees through the process with digital work instructions, while also carrying out in-line checks to ensure that every mouse was being built to spec.

Using a camera mounted at the top of the workstation and Tulip’s native machine vision features, the app running each bench would wait until an attendee approached and scanned their convention badge before beginning to display instructions on screen. Throughout the subsequent steps of the build process, users were guided by interactive work instructions while the quality of their work was verified by machine vision:

• The process began with attendees first retrieving the base plate of the mouse assembly, then the PCB which contained all the mouse’s electronics, and finally the optical lens. To help our operators locate each of these components from among the multiple parts bins, we integrated a pick-to-light system that illuminated an LED strip above the correct bin for each component. Our machine vision cameras monitored the process to ensure that the right parts were in fact picked.

• With those three components ready at the workstation, attendees could go on to the next step of beginning the actual assembly of their mouse. Thanks to the rich, visual work instructions built in the Tulip app, they were guided through the process of fitting together the lens and mouse base, followed by snap-fitting the PCB into place and press-fitting the battery contacts. Each time they were ready to proceed to the next step, attendees could advance the instructions using a foot pedal button connected to the Tulip app via an Edge IO device.

• With the base of the mouse ready, the app then presented attendees with the choice of three materials that they could select for the top half of their mouse. These included the unfinished SLS, tumbled and dyed SLS, as well as BioMed Clear resin with the various surface patterns. This step of the app again integrated with the machine vision and pick-to-light system in order to indicate which bins of parts were available to choose from based on current inventory levels. The app also recorded which material the operator ended up choosing.

• When the app detected that a mouse top had been picked by the operator, it then prompted them to identify which pattern they had chosen by placing the part under a machine vision-enabled inspection camera. The analysis results were displayed to the attendee showing which pattern the model had identified, as well as its confidence level. This information was again stored in the production record for that mouse.

• With the pattern validated, the app automatically moved onto the next step in the work instructions, which provided visuals guiding the attendee through the process of snap-fitting together the mouse top they had selected with the base that they had already assembled. Again, operators used the foot pedal connected to the app in order to advance once they had completed the step.

• With the mouse finally complete, all that remained was a quality-control step. Once again leveraging the pick-to-light system, the app directed attendees to take a battery from the inventory bins and install it to power on the mouse, as well as pick a bluetooth USB receiver in order to connect it to a computer. For operators to finish the quality-control step, they then had to connect their receiver to a test USB dongle and click on a target area displayed in the work instructions app in order to confirm its functionality.

• With the final quality control step completed, the app then presented attendees with a visual representation of their mouse's production record. Because we already had each attendee confirm their identity with an image of their badge in the initial step, we were able to extract their names using optical character recognition (OCR). We also measured the time it took them to complete the production process, and added these data points to their production record. We displayed all this data to them alongside a real-time comparison against other attendees, so that they could see how they stacked up against each other.

Having fully assembled and quality-controlled their device, attendees were now free to take home their new wireless mouse confident that it would function as expected. When all was said and done, the data we collected in Tulip showed that nearly 300 people had made their way through the Pop-Up Factory, with an average cycle time of 4 minutes and 55 seconds.

However, guiding attendees through the assembly of their mouse was just one part of the production process. Using Tulip, we built a suite of interconnected apps that were used to gain visibility into the process and make sure that the Pop-Up Factory was running efficiently.

One of the primary benefits of using Tulip apps to drive your frontline operations is that you can achieve real-time visibility into the status of your production process. Tulip offers powerful tools for creating custom data models and intuitive analytics, all without having to write your own code.

We leveraged these features in the second app that we built to support the Pop-Up Factory’s operations: a production dashboard displayed on a set of large monitors next to the assembly line. The app helped managers and operators gain insight into how efficiently the Pop-Up Factory was running thanks to the real-time charts and analytics. One half of the dashboard displayed process data, including a breakdown of the most popular materials attendees chose for their mouse, and an overview of the different amount of mice produced at each workstation. The other half of the dashboard provided an overview of the factory’s inventory levels, including a breakdown by workstation, as well as a historical view of inventory over time.

We also made use of this prominently displayed dashboard to help direct attendees to open workbenches. To achieve this, we used another powerful part of the Tulip platform — Automations — to track the status of each of the workstations. Automations enable you to run logic in the background of your manufacturing operations to make your workflows more efficient, accurate, and productive. By using conditional logic running in the background to determine whether or not each station was in use, we were able to then display this information on the dashboard and guide attendees to the right station. This helped ensure that our production process was running smoothly throughout the entire day.

Since Tulip’s software is cloud-based, running it on a variety of different devices or operating systems is a straightforward process. This includes everything from desktop, to mobile, and even edge devices.

We took advantage of this by building an inventory management app, which staff managing the production line could access from their phone — no matter where they were located in the Pop-Up Factory. This gave them real-time insight into inventory levels of various components at each station on the production line. When a particular component fell below the minimum-quantity threshold, the Tulip app triggered an alert, sending a message to managers using the mobile app that directed them to replenish the inventory with a fresh bin of parts.

As we were able to demonstrate with our Pop-Up Factory, a suite of just three connected apps developed in a single work week’s worth of time can seamlessly run frontline operations from end to end. Thanks to Tulip’s open architecture, integrating with everything from andon and pick-to-light systems, to machine learning and advanced analytics, is straightforward. This open ecosystem approach gives you the freedom to build your manufacturing tech stack around a wide variety of hardware and software solutions without being locked into a single vendor, which ensures that your business can remain adaptable.

While it may seem unrealistic to digitize a full-scale production line in just 40 hours, in today's fast-paced economic environment businesses need to be able to adapt quickly to changing market conditions in order to remain competitive. By embracing an open technology ecosystem you can develop a tech stack that meets your organization’s unique needs — unlocking new levels of efficiency, quality, and profitability.

We showed that with a platform like Tulip — which empowers citizen developers to create composable apps and leverage a wide range of technologies from many different vendors — you can build powerful solutions that can be rapidly implemented to drive positive change in your operations. By building your manufacturing tech stack on an ecosystem of technologies like those that helped us achieve success with this year’s Pop-Up Factory, you too can enable your organization to:

Empower Process Experts With Powerful Tools to Solve Problems

With Tulip’s no-code platform, citizen developers are able to solve complex production problems without needing extensive experience in programming or using complex automation software. Tulip’s App Editor provides an intuitive canvas for building apps with a drag-and-drop interface. You can easily create operator-facing interfaces, build trigger-based logic, and leverage custom widgets to drive your production workflows. This enables citizen developers to create and deploy solutions to the shop floor faster than ever before.

Make Operators’ Jobs Easier With Composable Apps

Operator-facing apps built in Tulip give your frontline workers the right information at the right time, and integrate with the technologies they use every day in their operations. With intuitive, highly visual, and interactive work instructions that integrate with IIoT devices like machine-vision cameras, you can ensure that your operators get it right the first time. Tulip’s unparalleled flexibility and ease of development also enable you to quickly incorporate operator feedback in order to continuously improve the experience of frontline workers.

Gain Real-Time Visibility into Every Aspect of Your Operations

Despite the fact that our assembly process was entirely manual, we were still able to capture a wide range of data about each mouse that was produced and calculate metrics like average cycle time. With operator-friendly interfaces that make it easy to collect data from the people, machines, and sensors throughout your business, you too can achieve real-time visibility into how your operations are running. Tulip’s powerful no-code analytics make it easy to build dashboards, track the KPI’s that matter to you, and react quickly to solve issues in real-time. You can set thresholds and build logic to trigger alerts via email, SMS, Andon lights, and more.