Interface for Robotic Arm
An intuitive user interface of a six-axis robotic arm prioritizing ease of use.
Challenge
To design a user interface for a six-axis robotic arm that helps the users connect and control robots easily.
Outcome
A high-fidelity Figma prototype with a design system prepared for seamless handoff to the developer.
Designed For
Orangewood Labs
Impact
-
The user interface complimented the already-designed robotic arm and helped the company pitch the product and raise money from prospective investors.
-
The project also established guidelines and design systems for future projects.
Collaborators
Design Peer, Design Manager, Front-End Developer, Head of AI and platform
Role
Research | User Flow | Wireframing | Design Assets |
UX Testing | Content Auditing | Aligning Stakeholders
Year
2022


About Orangewood Labs
Orangewood Labs is a Y-Combinator-backed startup building an affordable 6-axis robotic arm for SMEs. They leverage the latest advancements in ML to design and build highly efficient collaborative robots that are easy to operate.
Goal : To Democratize Robots
They aim to make robots accessible, strengthening the hands of every maker, turbocharging productivity at factory floors around the world or kitchens at home, and setting humans free from manual tasks that are monotonous, repetitive, or hazardous.


But...Programming robots is extremely complicated.
In the realm of robotics, various manufacturers have introduced their unique methods for programming robots. However, these existing interfaces often pose challenges, demanding extensive training and expertise from operators.
Brief
To design a user-centric interface for a six-axis robotic arm, such that it is easier for anyone to program and train a collaborative robot (cobot).
Yes, a huge task. But not impossible. Our objective was not to conquer it in one go, but rather to defy convention and embark on a journey to craft something that defies all expectations, something that nobody could have ever imagined.
Automation 101
Before identifying the opportunities and problems, we need to understand the different components of an automation rig. A diagramming explaining the basic functioning of a robotic arm setup is shown below.

Agile Methodology
As we were relatively new to industrial automation, we embraced an agile approach during development. This allowed us to complete multiple quick cycles of integrating and testing various features while learning more and more about automation.

Studying Existing Products
Before coming up with our own design ideas, it was important for us to understand what exists so as to save the effort of reinventing the wheel. However, access to the interface of such expensive robots is not easy. The interfaces, usually, are not open and require one to purchase the robot before getting access.
Luckily, I could access a cobot from UR Robotics at Orangewood and also several other interfaces from companies like Yaskawa, Svaya and more at IMTEX 2023 (Asia's biggest exhibition on manufacturing)
Hands-on experience of Universal Robot's teach pendant
I mapped out the screen flows of basic functions to understand what is essential and what can be eliminated.

Screen flow of Universal Robot interface





This helped us gain valuable insights into established design principles, user expectations, and industry standards. Some of these are given below.
Key Insights from Product Study
-
Stuck with legacy interfaces: The graphics and the overall interface, in general, felt dated, using symbols and terminologies that might be unclear to a new user. The primary reason behind the lack of updates was the longer training period that would be required after the update.
-
Lack of remote connection: Existing products don't have the ability to allow users to connect with robots remotely.
-
Lack of customization for different users: The robots are accessed by different users with different levels of knowledge and understanding, like programmers, maintenance teams, and shop floor workers. However, all of them were presented with a whole lot of options that might be irrelevant to the nature of their work. This often leads to longer training periods for the users.
-
Difficulty in visualizing the robot movements: The visual representation of the cobot's actions and state was unclear because of the possible lack of 3D visualization.
-
Availability of physical keys - It was noticed that interfaces with optional physical keys provided users with more usability and confidence, although it limited the scope of customization options.
Primary Task Flow
All of the market research that we did helped us understand various tasks and the flows that a user has to follow while automating a process. However, owing to the limited time in hand, we limited our scope and focussed on a few primary tasks, as shown below.
One of the major additions to the flow was the feature of remote connection, which appeared as two steps in the primary task flow - searching the bot and connecting to the bot.

(New Feature)
(New Feature)
User Flow
Based on the research done and understanding of the requirements, an initial user flow was created. It was never fixed and was always evolving and improved with continuous testing and consultation with experts.

Ideation through wireframing
An extensive amount of wireframes were created to test out different layouts, modes of controls, space division, sizes of elements, and various other functionalities.





(Click arrows to slide)

Limiting the Scope
As we explored ideas, the list of features to be included grew enormously, making it difficult to track and implement features in the prototype. We decided to take a step back and decide on the essential features to be included in the first revision of the prototype (MVP). I quickly did a content auditing, listing all the features on each screen. After having a round of discussion, we selected the most essential features to be included in the MVP.
While all the explorations helped us look at the bigger eventual goal, limiting the scope helped us realign ourselves and proceed efficiently.
Hi-Fi Screens
Sign-up/Log-in Screen: Orangewood's system has remote access to robots and hence different users can log in to access and connect to robots from any of their devices without needing a wired device connected to the robot.
Link a new bot: Users will have access to the only robots assigned by the admin. They would require a pin to link it to their account.
Connecting to a linked bot: Users will have to connect to a linked bot by entering the pin whenever they log in back to their account.
Using a template to add a new program: Users can choose a frequently used template program to add a new program efficiently and then edit it according to their needs.
Adding a new blank program: Users can also start afresh on a blank canvas and create a new program according to their needs. They can choose the tool/s that the program requires and even add description to it.
Creating a program (Add a new node): To create a program, users can add different nodes; like move, wait, tool state, etc.
Creating a program (Edit the node): After adding a node, the user can set the value to it either using Teach Mode (moving the robot physically with hands) or Joints Mode (input specific value to each joint)
Run a program: Users can log in back to run any of the saved program on the robot. They can also view different stats related to the program while it is running.
Dark or Light...
your choice

Testing 1,2,3...check
Total Participants - 6
Approach - Qualitative testing, interviews
Tools - Google Meet, OBS Studio
The high-fidelity prototype was continuously tested by industry experts and people who have worked with robots for a long time. Participants were asked to perform a list of tasks, and their actions were observed to identify further improvements required in the design.




Observations from testing
-
Everyone was happy with the look and feel of the interface.
-
Confusion in the flow of adding steps (waypoints) in a new program.
-
Missing speed function in the run screen.
-
No indication of program auto-save.
-
Difficult to find 'Environment Mode'.
-
Terms like 'Presets,' 'Add Customs,' and 'Preview' were unclear.