Building a robot optimized for lab automation
Opentrons make open source, flexible, user-friendly robots for life scientists. Opentrons robots automate experiments that would otherwise be done by hand, allowing scientists to spend more time pursuing answers to some of the 21st century’s most important questions.
In August 2023, the brand new Opentrons Flex was launched. It is a product unlike anything else on the lab automation market for a number of reasons. The Flex is fully open source, affordable, accessible, future-facing and has excellent UX. If you are interested, you can learn more about the Opentrons Flex here:
It was my responsibility to ensure the successful launch of the Flex from a software design perspective. This included the design and usability of the Opentrons App, the Flex On Device Display (Touchscreen) and various supporting software tools in the Opentrons Ecosystem.
Scientists have accepted that digital experiences in life science automation are sterile and cold (For a lot of robots, imagine Windows 95). Opentrons has the opportunity to deliver a world-class friendly digital experience that scientists want to be a part of.
E-commerce, social media, banking and productivity tools have gone through digital transformations to establish flexible patterns, implement design systems and pioneer best practices. As designers, we can leverage these learnings and apply it in the automation industry.
Product development is most successful when Product, Engineering and Design have equal footing to solve problems that balance product strategy, technical feasibility, and a great user experience.
Project discovery
The flagship liquid handler robot for Opentrons was the OT-2 and it was (still is) extremely popular for those looking to begin lab automation. They have been sold all over the world and users love them for their affordability, accessibility and ease of usability. Along with a hugely successful product allows you to learn, analyze and take the vision to a new level for the next gen robot.
The Flex has a lot of new features and hardware that will impact the users experience with the software. The more notable features include an on-robot Touchscreen Display (previously customers used a desktop computer app to control the OT-2), Gripper, LED operation light, 96 Channel pipette and deck customization.
As a team, there was an agreed understanding of what exactly was and wasn’t in scope for the MVP launch of the Flex. It was also clear what features will be fast-follows and what features will be on the roadmap for 2024.
Research
During customer research and interviews, these were the more important qualities that our customers required for a next gen robot.
- Increased efficiency, consistency and reproducibility of laboratory workflows.
- Anyone can install, setup and run their first protocol in less than an hour.
- Any scientist can adopt automation as a tool without spending expensive time upfront learning how to use it.
- The Flex fits on any lab bench
- The Flex can fit in any lab’s budget.
- Higher system reliability and more efficient deck slot usage means more walk away time.
- New instruments (96 channel and gripper) allow scientists to scale up their experiments in less time than they can perform by hand.
- Seamlessly enable third party integrations so that scientists can adapt the Flex to their lab.
Wireframes
There were a number of tasks the design team were working on over this period of time to allow the introduction of the Flex to our software. This included many feature upgrades to the Opentrons desktop app and protocol development tools like the protocol library and protocol designer. The main focus though was on the Flex touchscreen display.
Image: Example wireframe journey for first-time setup of Flex via touchscreen display.
Opentrons customers have historically used the Opentrons desktop app to operate the OT-2. Customers can still use the app if they wish, however the introduction of the touchscreen display enables users to operate protocols directly on the robot. This is a completely new and different experience to what our customers are used to.
User testing
I facilitated usability testing of the mid-fi design prototypes that Mel, the Senior Product Designer on our team designed. The purpose of the testing at this stage was to evaluate the Touchscreen Display’s user experience with real users to ensure that we are creating human-centric products.
Specifically testing the navigation, touch behaviours, gestures, new features, ergonomic position of the display on the robot, comprehension of content, identify roadblocks and uncover opportunities for improvement.
I am a big advocate of user-testing as soon as possible in the design process so that pain points can be discovered early and design hypothesis’ can be validated before any development begins. Generally speaking, the testing went smoothly, with these three things being the biggest takeaways.
Adding a back button to the top left of the screen helped users returned to previous pages rather than relying on gestures
Because there is a space constraint on the UI, we testing using gestures and all participants encountered some issue with gesturing.
Sending protocols from the Opentrons desktop app to the Flex touchscreen was a new concept users weren’t familiar with.
Further testing would be required when improvements to this functionality are made.
More reassuring user feedback was required when certain interactions were engaged.
Lab automation is an expensive operation so the UI and interactions should provide the user with the most support and confidence as possible.
Another key step is benchmarking this experience by asking participants to complete a third-party questionnaire so that the next time user testing is done, you can measure its success. There are lots of benchmarking tools available, but I use UEQ survey as I find it is straightforward to use.
Hi-fi designs
After many rounds of design, cross-functional teams and user feedback the hi-fi designs for the Flex on-robot touchscreen display were finalized.
Progress on the Flex had been moving along steadily and we managed to get our hands on a DVT (Design Validation Testing) robot for our first external customers to try it out and give us feedback. I was the lead facilitator across 7 beta testing sessions.
Since the software and hardware was about 60% complete, the main focus for the testing was simply the usability of the robot and setting up the robot to run a protocol. Some specifics included comprehension of an instruction manual, physically installing modules onto the Flex, installing pipettes, installing the gripper and completing some protocol setup tasks using the software.
The beta testing was an incredible milestone for the team and the hard work was paying off because the feedback was mostly positive. The session left us with so many insights it helped pave the way for future product roadmap ideas. These were some of the major takeaways
- Module installation requires strong support material for first time use
- The DVT pipette calibration probe needed usability adjustments
- Identify and prioritize Opentrons app bugs
One example insight from the beta testing was that users had trouble attaching the pipette probe to the pipette for automated calibration. Based on the feedback provided, the hardware team designed and developed an improved iteration.
The feedback gathered from the beta testing sessions helped the product, hardware and software teams to prioritize improvements to the experience. Retrospectively speaking, the insights from the beta testing sessions were actioned and completed by the time the Flex launched.
Scope changes
There were some unexpected key requirements introduced later in the project which can disrupt delivery and timelines. The design, software and product teams worked together to provide the best possible user experience based on our internal knowledge. The drawback to squeezing in features close to launch means they haven’t been tested with customers.
To keep work moving forward, the design team used Loom to translate product requirements into design concepts. These videos proved crucial in helping communicate design decisions, seek approval and most importantly – free up calendars.
Design QA
In the late stages before launch, a lot of the robotics team focus was on testing. The importance of different teams testing a product means there are different perspectives and ways of seeing things. The design team was able to identify UI, interactive and copy differences between software and approved designs in Figma. This testing was crucial to ensure the final product is polished and behaving as expected.
Delivery
The first US customers receive Flex in August 2023. Since then it has been very successful as interest and sales consistently perform. If you want to find out more information, check out the Flex landing page on the Opentrons website.
Opentrons is working hard to integrate AI into protocol development, compatible with the OT-2 and Flex as well as many other features coming in 2024.
Acknowledgments
I am incredibly grateful to have had the opportunity to work on such a meaningful product to advance the world of life sciences and be part of such an inspiring community.
Thank you to the Design, Software, Product, Science, Hardware, Marketing, Support, Sales and Operations teams for all of your collaboration on delivering such a meaningful product to the market. I’m looking forward to seeing Opentrons success as they continue to quickly grow.
Other projects
Check out some of the other projects I have worked on below.
A platform for people to book unforgettable travel experiences. Explore the process of how decisions were made on this highly complex Flight Centre product offering.
A systematic approach to design that helps build cohesive experiences across products, applications and platforms.
How might we better communicate to our customers and prospective customers that digital payments are safe, secure and easy to use.