Expert Interview: Inkbit CEO, Davide Marini, on the Potential of Multi-Material Inkjet 3D Printing
10 December 2019
3D printing is becoming a more intelligent process, as more companies look to integrate Artificial Intelligence (AI) into the technology. One example is Inkbit, a US-based start-up, which has developed a multi-material inkjet 3D printing ‘with eyes and brains’.
Although multi-material 3D printing has been around for a while, the technology has been used primarily for prototyping purposes. Inkbit aims to revolutionise the technology by developing an inkjet 3D printer capable of end-part production.
To learn more about the new multi-material technology, we caught up with CEO of Inkbit, Davide Marini.
In the interview, Marini explains what makes Inkbit’s technology so unique, its key applications and also shares the company’s outlook for the year ahead.
Can you tell me about Inkbit?
I was introduced to the inventors of Inkbit’s technology while they were still working on their early prototype at MIT. At the time I knew very little about 3D printing, yet the idea of endowing a machine with a set of eyes immediately captured my imagination. We eventually spun Inkbit out of MIT in the Summer of 2017.
The key differentiating aspect of our technology is a vision system, integrated inside our 3D printer, that makes the machine intelligent.
As you know, 3D printing works layer by layer, but in our machine each layer is scanned at micron resolution immediately after deposition. If there are any deviations from the expected geometry, they are immediately corrected in real time by remapping the next layer.
This element takes care of all the random errors. For example, it could be a clogged nozzle in the printhead, or any type of error that is not predictable. The interesting aspect of having an integrated vision system is that, not only does it allow us to correct these random errors, but it also allows us to predict material behaviour during the print process.
And we do this because we have access to the data set from each scan of every layer. Let’s say, for example, that a material tends to shrink; because we scan every layer, the machine can learn the particular behaviour of that particular material. So the next time it will print a slightly larger geometry to compensate for shrinkage in advance.
It is because we have a vision system integrated into the machine, and have endowed the machine with a set of eyes, that we can now build specialized AI algorithms to pre-empt the systematic errors that may come from specific material behaviour, such as shrinkage, or flow, etc.
There is also another advantage: every part we print comes with a digital record. This is possible because we scan every layer so we can, essentially, reconstruct – almost like in a medical CT scan – each part at the end of the print. This enables customers to perform 100% quality control.
Let’s say, for example, that you want to print a very intricate fluid manifold, with an intricate structure of internal channels. How do you know that what you printed is actually what you wanted?
In our case we know, because the print was scanned at every layer. That’s a big advantage for the customer.
Lastly, thanks to our vision system, our printer manufactures parts in a purely contactless way. So there’s no need for a mechanical flattening device, which is today required by material jetting. And this allows us to print with better materials.
As can be seen, we have lots of advantages, but they all come down to one single principle, which is the idea of a vision system integrated into the machine.
How does your technology differ from other technologies that are currently available?
In terms of the benefits to the customer, there is not a single solution today, at least to the best of my knowledge, that allows for 3D printing of different materials in the same part, and with production-grade materials.
If you imagine a two by two matrix, where on the X-axis we put, say, single material versus multi-material, and on the Y-axis we put prototyping versus production – that two by two matrix is fully populated, with the exception of one box: multi-material for production.
Today, nobody plays in that space because there is no technology to do that. For example, Polyjet machines make stunningly beautiful parts, but these printers were designed for making prototypes, or parts that look and feel like the real products but cannot be used in the real world. This is mainly because the materials were not intended for that purpose; they’re not capable of withstanding the harsh treatment they would receive in a part, say for example, for a car.
Similarly, technologies like Multi Jet Fusion and FDM, can make parts with excellent mechanical properties, but they’re all single-material parts.
Our technology enables to harness the power of inkjet, a technology that has been around for many years, and use it into the world of production.
Our machine is designed for production, for making parts that contain, for example, both a soft and a rigid area, in the same build. Say, for example, you want to build an athletic running shoe that contains both rigid and soft parts in the same print. We want to be able to do that. In terms of applications, we’re looking at the medical field where you sometimes need multi-material parts.
In terms of benefits for the customer, we want to develop a multi-material platform, with production-grade materials and with the type of reliability and consistency that is needed for real high-volume production parts, and in a format that enables 100 per cent quality control.
Could you expand on the types of materials you produce and their benefits, in terms of applications?
At the moment, we have 3 materials, with plans to develop more in the future.
First, we have an epoxy that is a high-temperature-resistant material. And this can be used in applications like electronics, or in areas that require the movement and distribution of high-temperature fluids. Our material is a real epoxy, not a mixture of different chemistries.
The other 2 materials are rigid and elastomeric biocompatible materials. The elastomeric material is particularly interesting, because it has a very high elongation at break – about 800%.
Inkbit has recently announced a $12 million funding round. What does this mean for Inkbit going forward and how does it play into your future plans?
We decided to invite strategic investors to Inkbit, because we very much believe in partnerships. And I believe that in order to develop the best machines and to make the best technologies, especially in 3D printing, it requires expertise in so many different areas.
For example, to make an outstanding 3D printing machine for production, requires expertise in three different areas. It requires mastering hardware, chemistry, and cutting edge software, especially when we’re talking about AI.
So it’s just really, really difficult, especially for a start-up company, to master all 3, because it’s equivalent to starting 3 different companies.
I really like partnering with existing leading companies that are experts in their field. Together, we can bring to the world something truly spectacular. And so I’ve invited companies in all these areas.
We’ve two materials companies – DSM and 3M – some of the world’s leading materials companies, and Stratasys, who is the world’s leading 3D printing company, especially because they invented inkjet. So they are the world’s experts in inkjet technology.
We also have a British company, Ocado. The reason why we really like them is because they bring us specific applications in robotics.
What are some of the challenges you see when it comes to accelerating adoption of 3D printing?
The first thing, I would say, is materials. We don’t have materials, yet, that are at least equivalent to non-3D-printed materials, at least in the field of polymers. I would even venture to say that 3D printing should be able to offer better materials than the ones that are available today for injection molding, but there is still a long way to go. So materials is the number one challenge.
The second challenge is the reliability and accuracy of the machines. That is, making sure that the machines consistently make parts that are faithful to their 3D model and can operate continuously for long stretches of time.
And the third one, I would say, has to do with the mindset of the product designer, where engineers and product designers are still accustomed to thinking in terms of injection moulding, while 3D printing offers a much wider design space. It will take time to make people aware of the opportunities offered by 3D printing. But this is more of an opportunity than a challenge.
How are you going about addressing the challenge of changing people’s mindset?
The way we’re doing this, is by focusing on applications and developing both the materials and the machine in close collaboration with our customers.
So, as an example, we have been working with Johnson and Johnson on a product that they were manufacturing with injection moulding. But when they came to us, they invited us to offer our own input into the design itself, which required designing a new 3D printable material.
In collaboration with J&J, we have designed a specific material for this application. Since our system is very modular, we’ll also be able to design an entire machine that will make the product.
Consequently, our approach to the market will be very applications-driven. This means that we want to be close to the customer; we want to first know exactly what product the customer will want to create, and then we’ll design the entire work.
How do you see additive manufacturing evolving over the next 5 years?
I think AM has the potential to change the world. I also think that we are lucky to live in a time where, in front of our eyes, there is a revolution happening in the way products are made. I’m actually very intrigued by the idea of offering the power of a manufacturing line to everybody.
What will 2020 hold for Inkbit?
The most exciting aspect of the next year will be testing our machines at customer sites. So we will be building a few copies of our machine and it will be our alpha prototype. And we’re looking for early adopters. We want to select a few sites, a few customers that are interested in testing the machine at their site.
This will happen within the next 18 months, once we have completed the final design round of our current prototype. And so I would say, within 18 months from now, we want to have at least 5 partnerships for 5 beta installations. So, that’s going to be the most exciting aspect of taking our machine to the factory floor.
To learn more about Inkbit, visit: https://inkbit3d.com/