The difficulty in building great hardware

One of the things people ask is why can’t we hire an artist and build a nicer robot design. If only building hardware was as simple as that.

We had multiple artists working full time with us. We hired one of the best product design firms in the country. And they have helped us greatly improve things. But, it is not a simple question of hiring an artist and things will magically fall in place.

India is not short in artists, but how many world class products we have? Why can’t we make great designs given the lakhs of great artists around.

Because sketching is not the same as design.

Let’s say an artist is tired with the same old rectangular phone everyone builds. What if she tries to build a cubical phone? It would look cool, but would it be practical?

Or someone draws a circular space ship like a car. How feasible would be to reproduce that in production?

Here are some key challenges:

  1. Details: There are 18 major components and nearly 75 minor components in the robot. Getting each of them designed to work with the entire entity is complicated. A lot of artists having trouble getting into the level of details and building 3d modeling.
  2. Cost: Producing something commercially is a very, very expensive process. Especially a product of our size. A basic 3d printing would cost Rs. 5,00,000 to build just the outer shell of one robot 5 foot tall. How would we make that commercially viable? Building injection moulding – the best process for accuracy – will cost Rs. 8,00,00,000 just to make the 18 stainless steel mould parts needed to build the whole robot. Since we don’t have that kind of money, we have to resort to more inexpensive manufacturing techniques – where intricate design would not be replicated.
  3. Production process ability: To accomodate our manufacturing techniques – such as FRP moulding or vacuum forming – less expensive but still running into 10s of lakh rupees – we have to resort to simpler curves and make a few compromises. Our present prototypes use even cheaper techniques – completely built by hand – but has more limitations on the design.
  4. Stability: The larger you go, it is not just cost, but the mechanical complexity grows enormously. We have to worry about such a large moving robot toppling and falling down. That means we have to worry about the dynamics and do a few “ugly” things like having a large base to ensure the CG lies within the base always.
  5. Strength: And the large size also means we have to worry about the strength. A 3d printed object of that size [built through a lot of smaller components] and we have to accomdoate both the material and the design to ensure that strength so that the bot doesn’t bungle and fall apart in sheer stress.
  6. Customer needs: Planes would be safer and cheaper if they didn’t have windows to look out. But, that would be unacceptable to the customers. In our original design, we didn’t want to put a tablet as that ruins the experience. But, our customers insist on it and that means we have to insert a tablet somewhere. A lot of practical, customer requirements means you have to make compromises. Business is not art.
  7. Competing pressures: Body design is one among the 20 odd things on our place – ranging from software design to electronics design to sound design [making sure people can hear and talk to the robot in noisy conditions] to website design to brochures. That means we are always hands full and prioritize.
  8. Revenue pressure: Our engineers never the luxury of building things peacefully. As a bootstapped startup, we are always in a hurry – to get as many sales, revenues and event showcases. These are important for the healthy of the company and to get feedback to make sure we are running in the right direction. However, this also pushes the timelines when it comes to design.
  9. Time: It is easier to verify the answer or proof than to write the proof/solution. Often the issues show up only after the fact and not ahead of the fact. This requires a lot and lot of iterations. You can see the iterations in the past 11 months.

Look back in time to see our improvement.

Mitra robot

MITRA – 1st Prototype

MITRA robot


MITRA robot at birthday parties

As we go along, we will keep improving. And when we keep making new designs, the older designs would look ugly and funny. Perfection is a process, not a state.

By | 2018-11-26T11:56:41+00:00 June 2nd, 2018|blog|Comments Off on The difficulty in building great hardware

About the Author:

I'm the cofounder and CEO of Invento Robotics. I have been in tech industry for 12 years and have worked in a range of products starting from Microsoft Windows in Redmond. I'm also the most followed writer on Quora and a winner of multiple international awards for research and innovation.