Did you know humans have over 2000 touch receptors in their finger pads to help them grab, touch, grip, and recognise various objects? Yeah, neither did we. Turns out, robots need a little sensory magic too. Enter Contactile, led by the brilliant engineer and CEO Heba Khamis, ready to give robots a much-needed touch-up.
This interview originally appeared on Missing Perspectives by Natasha Gillezeau and has been republished with permission.
In a beautiful building in Eveleigh, Sydney, science and technology nerds will find a deep tech incubator called Cicada Innovations.
Inside, Cicada is abuzz with various startups at the cutting edge of deep tech trying to take what researchers have learned at universities to the next level of commercialisation.
Contactile is one such startup. Led by engineer and CEO Heba Khamis, the Contactile team are developing grippers and sensors to help robots integrate a sense of touch into their activities.
Human beings naturally have about 2000 of these touch receptors in their finger pads to help them grab, touch, grip, and recognise various objects. To do the very same tasks, robots struggle a bit (okay, maybe a lot), which is why a sensor containing nine sensing elements developed by the Contactile team could make a huge difference to the field.
Beyond the core technology, Heba also wants to do innovation differently. Steve Jobs is hailed as an icon for his work at Apple, but Heba bristles at some of the tech company’s approaches to doing business.
“The thing that bothers me is I don’t like innovation that ‘blocks’ collaboration,” she says. “I think Apple is really innovative, but I don’t know if it’s the commercial decisions on the App Store, or how they approach ports and connectors, I find that that’s inherently an obstacle to collaborating.”
Khamis continues, "I also think that I conflate being an innovator with being a good leader, so empowering the team is a must, and there are a few extremely innovative leaders out there who may not be doing a great job at keeping their team happy."
Meet Heba Khamis - a founder who has made the leap from researcher and lecturer at UNSW to full time robotics CEO - and is building her company her way.
Natasha Gillezeau: Hey Heba. Thank you for making the time to chat to me today. So, one of our MP contributors Cara Davies - who started and sold a Gen Z fitness app - recommended a book called The Unfair Advantage. As an entrepreneur, what would you say are your unfair advantages?
Heba Khamis: I want to say naivety. I think it does take a bit of naivety and not knowing what you’re getting you’re getting yourself into to become an entrepreneur in the first place.
I don’t know if you know the quote “we didn’t do it because it was easy, we did it because we thought it would be easy”. There are a lot of unknowns, and you just have to be comfortable in that, and that does take a bit of naivety.
You can only do what you can do, and there are a lot of things out of your control and once you’re comfortable in that unknown, that’s when you can focus on doing the things that you do have control over and progress the startup as best you can in an unknown, unpredictable environment.
Is there anything else that comes to mind in terms of your education, or where you grew up, or anything else that you think, oh, these are some of the surrounding circumstances?
Everything contributes in some way. I grew up in a household where there were no limits. In my family there was never “this is a girl job, this is a boy job.” It was really about doing what you wanted to do as long as it’s going to make you happy, and in some respects, keep you financially stable.
Everything from the way I was brought up, to my engineering background, which is good for entrepreneurship because you’ve got that part of your brain switched on whether it’s an engineering technical problem, or a business problem, or a customer problem.
The universe brought together our founding team, and that’s been a key factor in our success so far. I’ve seen other founders who’ve struggled to find co-founders or that core team. If you don’t have that, you’re stuck.
Let’s start with some basics and we’ll do a bit of ELI5 ("explain it like I’m five years old"). What exactly are you building at Contactile?
Basically, if I put something in front of you, and I asked you to close your eyes, reach out, and pick it up, you can do that.
You don’t actually rely on your eyes to be able to hold something. What you rely on is your brain, but you rely on your sense of touch. You’re holding a cup of coffee right now - you’re not looking at it, you’re looking at me.
So there’s no vision involved in this process - vision helps you reach for the item, but as soon as you touch it you’re relying on your sense of touch.
Now, robots have traditionally not had any kind of sense of touch, and that limits what they’re able to do, because they need to be pre-programmed for very tightly constrained environments.
If a robot needs to pick something up, they need to know exactly where it is, they need to know how much pressure to apply to that object in order for it not to slip from their grasp, and they need to know all the properties around weight, size, mass and distribution. That constrains how we can use robots because the environment needs to be well understood, and predictable.
So what we do is we make the sensors that go on the hands of the robot that enable the robot to handle anything it comes across without knowing anything in advance about that object.
How are you building the sensors? How does that work?
We started at UNSW in Sydney at the Graduate School of Biomedical Engineering, and the project that inspired this was a project in human tactile neurophysiology, and we were looking at how do humans sense all the things that we need to sense in order to use our hands the way that we do. And that project is what inspired us to start developing the tactile sensors for robots.
There was a lot of build up to ‘will it work?’ before we even spun out the company. Once we had that under our belt, it became more of a manufacturing challenge. We now have a lot of the parts of the sensors outsourced, and then we bring those parts together, and assemble them in house, and we calibrate the sensors in house, so the value-add is here in Sydney.
We make the sensors - there is still a lot of development in terms of optimisation and potentially miniaturisation. To give you context of scale, there are about 2000 touch receptors in the human finger pad. Our sensor at the moment, we’ve reduced the complexity, and we’ve got nine sensing elements in our sensor.
What has building this company taught you about the distinction between the human the machine? Because you’re bringing something from the human - the senses - into the machine.
You need to choose your battles, and which hill you’re willing to die on, because if you try to actually replicate the functionality of a person, you’re going to be at it for fifty years, that’s not a short term problem.
So you need to focus on what technology do I have at the moment, and what are the customer problems I can solve as is, and then start to look at okay, if I only spend another six months solving one aspect of this, what new customer problems can I solve with that.
It’s an iterative approach, and there’s a huge gap between what a human can do, and a robot can do. Things that are easy for robots are hard for humans, and things that are easy for humans are hard for robots. There’s a fundamental difference in the way that the two things are set up and how they operate.
We’re trying to close that gap, but you can only do that step by step. We’re not going to have a solution tomorrow that’s like here’s something that can actually function as well as a human in these very unpredictable environments.
What’s one thing outside touch and grip that’s easy for humans?
An example is when you know that the garbage truck is coming the next morning, and you’re like “oh crap, I haven’t taken out my garbage and you’re trying to put the key into the lock, and it’s dark”, but somehow you manage?
A robot, even if it could see it, that’s an extremely challenging tactile task. It’s one example of a highly tactile task. To position the key based on purely vision - you need to be super, super accurate and precise. But if you have that sensation of touch you can angle it, and feel when it’s made its way into the hole, and then you can feel when it’s smoothly going in or when it’s hit the end. All that stuff you can feel, you can’t see very easily.
So that’s something that’s relatively easy for humans (thanks to years of learning and training), and hard for robots.
What skill-set did you bring to the table when you started the company?
I have an engineering background. My area was biomedical and signal processing and software. We do have a little bit of that in the company, but most of it was project management. I’d been in research by that point for eight years, managing projects, managing teams, that’s my strength, and then the rest they call it “drinking from a firehose”, which means having to upskill very, very quickly in a lot things you had no idea about prior.
On the “drinking” piece, what’s something that you’ve picked up, even if it’s a small thing like website design?
We’re a small team, so as CEO I wear many hats. So I have to talk to customers, business development strategy, financials, accounting, legal, marketing, sales, social media, and website. List everything that isn’t a technical function in the company - and I wear that hat. I still take out the garbage on occasion - it’s literally everything.
Do you enjoy being in a role where you’re across a lot? Or do you find the context switching quite difficult?
The context switching is challenging - even if it’s the same role. It is a drain. The challenge of wearing so many hats is it’s constantly fighting the priorities. Figuring out what is a high priority and what isn’t when pretty much everything is a high priority. For example, before this call, I was renewing our insurance. It’s a tedious, menial thing to fill out a renewal form, but I can't ignore it.
The boring stuff allows the fun stuff to happen!
If you were to boil down the role of the CEO to one sentence, it would probably be enabling everybody else to work on the important stuff so that the company can thrive.
Who are your customers? And how has that evolved over time?
In the last 12 months, there has been a proliferation of companies trying to build humanoid robots, and we know for a fact they need tactile sensors. They probably don’t need them right now, because they’re going to be doing simple tasks to start with, but we as a company won’t survive.
In terms of customers today, we do sell a bit to researchers, whether that’s in academia, universities or larger companies, we’re looking at some industrial customers potentially in automotive, and other types of manufacturing.
What role does Cicada Innovations play in your life?
We won a prize that got us incubation at Cicada, we got one-on-one mentoring here to help us think about customer discovery, business development, and how to raise money. So we’d already been through that when we did raise our seed round. And it was like, now we can afford to rent a space.
Learn more about the Fast Start program here.
How much did you raise?
US$2.5 million from a US investor. It was like - finally we can breathe - we need space, desks, labs, all the fun toys so we can get on with building sensors. And when we looked at renting an empty office versus being in Cicada as an incubator - it was a no brainer.
Empty warehouse means we’re completely isolated, being here means I can go next door and say,” hey, I see you’re going through a similar problem, can you spend 15 minutes with me talking through this?” It’s the most powerful thing being able to talk to other founders and even just being visible.
There’s another company here at Cicada who has a very similar technology, and they’re about 4-5 years ahead of us. Even just being able to see that, the size of their team, the makeup of their team, how they’re doing their manufacturing, what to raise when - even just having that in front of you is so powerful.
What an incredible resource, and the collisions that can happen.
So the problem is - I usually only leave the office to go to the bathroom - and I often run into people on the way to the bathroom. The timing is hilarious.
Have there been any interesting incidental connections that have happened?
We talk to each other a lot about challenges, I might run into someone who is struggling with writing a grant. And it’s like - I just wrote one - do you want mine for reference? Stuff like that. Or I’ll be walking around with bits of cardboard - and they’ll be like, what are you doing Heba? Working on packaging. Ooh - I know someone who can help you with that.
What’s your aspirations for Contactile in the next 3-5 years?
One of the big things for us at the moment is customer traction - figure out if there is a particular industry or use case where it makes sense for our sensors or grippers.
The challenge that we’ve had in the past is we like to talk about general purpose, but general purpose is hard to sell, because people don’t necessarily make the connection between what their robots are doing, and what we're offering.
So if I say I have a general purpose hand, they might say, "well I need something that picks up this part, moves it over here, reorients it, and moves it over here." That is general purpose, but they’re not necessarily understanding the connection.
Is that a branding issue?
It’s also an education issue. At Contactile, we’re just adding another tool to the kit that allows a robot to do new things. But because it’s a new tool, people are still a bit unsure.
It seems like an interesting time for the company, particularly being involved with such frontier technology. Thank you so much for your time Heba and good luck!