Robots in the Wild: Moxi to the Rescue!
Diligent Robotics is building robots to assist hospital staff with routine activities so they can focus on caring for patients. Moxi, Diligent’s autonomous mobile robot, supports understaffed and burdened healthcare staff by doing deliveries 24x7 throughout a hospital or care facility. The company is now deployed in several hospitals and recently raised a $30M Series B round lead by Tiger Global.
And it couldn’t come at a better time. The industry is facing massive and unprecedented labor challenges, with nearly 1.8 million unmet job openings as of April 2022, nearly 16% of the total. Labor shortages were the top concern for hospitals in 2021, displacing financial concerns as the top priority for 16 years running. The American Hospital Association recently testified to the US Senate describing the issue as “critical”.
Grit is proud to support the mission. More from Vivian Chu, the co-founder and CTO of Diligent Robotics, below.
Vivian Chu is the Co-Founder and CTO of Diligent Robotics
JGR: Hi Vivian! Tell us a little bit about your background and how you got interested in entrepreneurship?
VC: Hi Jennifer! My parents were Silicon Valley software engineers who encouraged me to study engineering. My Mom first introduced me engineering at Hewlett Packard on “bring your kid to work day.” My Dad eventually moved to China to found a software company that ultimately failed, so I got to see how hard entrepreneurship can be. But I saw how he enjoyed the autonomy and that inspired me.
JGR: I also grew up in Silicon Valley and had a father who was an entrepreneur. I thought about getting a PhD but ultimately stopped at my Masters. How did you decide to get a PhD? Tell me about your academic journey?
VC: Like most engineers, as a kid, I liked to tinker, take apart toys, and break things, but not necessarily put them back together properly! I studied computer science at UC Berkeley. It was the first time I encountered programming.
I never truly connected to the work in my computer science classes until I took my Intro to Robotics course junior year. They gave us a Roomba iCreate platform and I saw how programming could interact with the physical world. We mapped and explored rooms with the Roomba and even got the robot to climb a ramp using its accelerometer onboard.
I had a great mentor at Berkeley who told me I simply missed the physical connection to my engineering program and encouraged me to continue studying and consider research.
Before I did a two year pre-doctoral program at IBM, I always thought PhDs were for the geniuses. But the researchers there convinced me that a PhD just required hard work and deep interest.
I still couldn’t quite make the leap to a PhD and went to Penn for my masters. There I learned about haptics and how robots could interact with the world using various behaviors and sensing. I programmed a robot to learn about haptic adjectives like soft, hard, fluffy, by having the robot squeeze and interact with different objects, but it took me months to just get the robot to do these basic motions and I thought there has to be a better way. Then I heard about Dr. Andrea Thomaz at Georgia Tech doing really cool work by programming through demonstration and I joined her PhD lab.
Dr. Thomaz went to the University of Texas my third year of my PhD, so I spent the summer setting up her UT lab and it was first introduction to Austin, TX. We decided I should finish my degree at Georgia Tech but continued to work with her at UT.
As I was finishing my PhD, I was considered starting an academic career, building a lab from scratch or starting a company directly. Andrea convinced me that founding Diligent Robotics would be just like a two year post-doc and to give it a try, and here I am four years later!
JGR How did you and Andrea do customer discovery and land on the hospital supply delivery opportunity?
VC: Andrea a NSF i-Corps grant to explore robotic technology. Then Andrea secured an SBIR phase I grant focused on healthcare robotic applications and I joined Diligent Robotics as a co-founder. Along the way, I took a break from my PhD for 6 months, went to Austin, coded away the initial codebase for Diligent in a WeWork office and shadowed hospital customers. We figured out quickly that there’s a pain point around fetching hospital supplies.
We validated our i-Corps hospital worker interviews by shadowing staff with a stopwatch and a clipboard. We brought our robot to two hospitals in Austin, Seton and St. David’s for several days of actual demos and testing. We supplemented the experiments with simulation software to truly explore what was possible.
JGR: You could see the extreme labor shortages and that the opportunity was big, what made you think you could solve it?
VC: Hospitals are very complex systems. We saw how nursing is impacted when hospital systems break down. I remember watching a care tech literally running from nursing station to nursing station to locate and change out linens for a particular patient room. I also saw a nurse nearly almost dispense the wrong medication due to distractions. It was surprising to see them stashing wheelchairs, in the corners, to prevent delays on patient discharge. Our motto became, “you are the best of human kind. How can we make your life easier.” It wasn’t just a labor issue, but an opportunity to ease their cognitive load. Labor shortages may drive nursing burnout, but we saw how other burn out factors are clearly not a good use of their time.
JGR: What was the biggest technical challenge you needed to solve in getting to a working prototype?
VC: The hardest challenge is not developing a specific functionality. Instead, it’s achieving robustness and reliability. It’s definitely a technical challenge to get from prototype to feature complete. But getting to production ready is a higher bar. If you see it glitch once in the lab it will glitch a hundred times in one month at the customer site. We worked hard and are continuing to work hard to create a truly robust platform.
JGR: What about navigation challenges? Isn’t it hard to teach a robot to navigate a hospital autonomously?
VC: Yes, navigating indoor environments is a non-intuitive challenge. Sidewalk robots and self-driving cars out in the world can rely on pure LTE anddirect GPS. WiFi in a constrained indoor constrained environment is OK, but not all hospitals are perfect. We had to improve our network site assessments and addressed how to fail over to LTE amidst WiFi dead spots.
JGR: What happened when Covid hit? How was the company impacted?
VC: That was a scary, but lucky time. We had just closed our Series-A a week before Covid hit. We then threw our S-A plan out the window and cocooned by focusing on research with our early customer. By mid 2021, most hospitals had managed multiple waves of Covid, but staff burnout and labor shortages only worsened and so they needed a long term solution. Our Moxi platform wasn’t just an interesting innovation but something they desperately needed today. We went from a “nice to have” to a “need to have” solution.
JGR: Did the use cases change?
VC: When Covid hit, we considered UV cleaning. But customers said no, no, keep your focus on transport. We did small sprints around other use cases, but realized that doing deliveries was still what was needed most.
JGR: Any feedback from patients, non-staff? How is Moxi perceived?
VC: We get a wide range of reactions. Usually it’s “Did I just see a robot?” and then they ask “Oh, the robots are here to take my job?” We’d reply “No, we only want to take the annoying parts of your job away and we draw the line at the patient’s room. We want you to do the patient care.” Then we get the aha moment and they start throwing out ideas about what the robot could do for them too. Then the reaction I really love is after the initial deployment period, weeks and months in, customers tell us, how useful/helpful robot is, that it reduces steps and that Moxi brings a little joy to their lives. Moxi is super adorable.
Here’s a great story from the VP of Innovation at a customer hospital while rounding with his team at a nursing station. The nurse manager, saw the Moxi, paused the meeting and said “I love that robot” with a big smile on her face.
JGR: Does Moxi have a voice?
VC: Moxi has prerecorded voice lines including “I’m here for a delivery” “Hello” and “Thank you.” But we don’t want people to think that the robot is more intelligent that Moxi really is and we mostly rely on meeps and sounds to convey intent. Moxi is a teammate on the transport side not patient care.
JGR: Where do you think the company will be in 5 years? What needs to happen for Diligent to be a market leader?
VC: Before we launched Moxi, we learned a lot about what a general purpose robot could do beyond point-to-point transport in a hospital. For example, materials management in a warehouse or a grocery store. We are committed to serving hospitals though. In 5 years, I can see us having a second product line beyond deliveries and have more integrations into healthcare software systems. We think we’ll succeed in the hospital based on our deep domain knowledge and are currently committed to the space.
JGR: The composition of teams is different in all the verticals and given their unique use cases, your strategy makes a lot of sense.
JGR: Diligent is a venture-backed company. What is your experience with your venture investors?
VC: So far it has been pretty positive. We have a great set of investors, folks we can trust. They have deep industry and operational experience. They give us advice but let us make the right decisions given that we are in it day in and day out.
JGR: How do you describe the company culture? What are the values?
VC: One of our venture investors has a VP of culture. She said your values are reflected in the things you do and how you do them. Our values also are reflected in how we hire, onboard and provide ongoing education.
During our new employee onboarding sessions, Andrea and I talk about how the company got here with radical candor and psychological safety. We understand we are a startup that will make mistakes. We have a motto, “up and to the right.” Radical Candor is about how can we fix problems. Let’s all be on the same page. If we don’t tell each other what’s wrong, we’ll make the same mistakes again.
We are super mission driven to make a difference for nurses and clinicians. Half of our employees have a family member in health care.
JGR: Do you see yourself as a role model?
VC: I don’t see myself as a role model, but I know it’s important. I try to attend events for women in STEM and LGBTQ events so that people can have role models.
JGR: You are too modest, you are a role model. Every time you show, you are inspiring somebody.
JGR: Anything you want to add?
VC: We are hiring! Join Diligent in transforming healthcare!