Beginning of a New Age
Exploring the future of robotics and AI, sharing insights from building a robotics product and the challenges of hardware development in India.
The last couple of months have felt like exploring a magical forest filled with enchanting sights & discoveries, but with hidden brambles and thorny paths at every turn.
We started the journey of creating an amazing robotics product, inspired by Boston Dynamics. With an ambitious vision, we toiled every day to piece the machine together, negotiating with the challenges on the way. While the journey to getting there was a bit disappointing, they also brought a lot of hope and faith in the journey we have embarked on.
We are genuinely at the beginning of a new age in science and technology, and it's one of the best times to start building something impactful if you're capable and believe in your skills.
With the spotlight on AI and significant investments being made in computing power, everything electronic and mechanical will have a chip slapped on it.
In the coming future (5-10 years from now), you will see new editions of all the current devices you use in daily life, which will actually be smart. No not "smartphone" kind smart, but smart enough to learn their strengths and weaknesses and calibrate themselves to maintain optimal efficiency for their intended goals. Just as humans adapt because our cells and organs adapt, each part of these machines will become smarter and adapt to produce the best results.
The question is not who will build the smartest machine (we are very far from singularity), but who will build the intelligence for each of the parts of that smart machine?
When it comes to robotics, we are in a similar era to the early days of personal computers. We are in the 80s and 90s of robotics, much like computers were back then.
There are a good number of companies building humanoid robots, some of which are truly exceptional. In the coming decade, just as people now own personal computers, they will own personal robots.
For this to happen, both the software and hardware aspects need to advance significantly and made affordable. We will see a handful of billion-dollar companies emerge in this space, providing intelligent parts and software services to the robotics industry.
Our journey began with designing a miniature prototype of our machine. It resembled a semi-humanoid, small moving cart with a robotic arm attached to the top, tasked with autonomously picking up miniature bricks and tiles to build walls and floors.
We chose to develop an initial version in a miniature form to reduce R&D costs and time, allowing us to focus on creating a reliable demo. This demo was intended to attract early customers, encouraging them to pre-order and invest in the machine's development.
This project involved creating two major components:
- The Moving Cart: This cart needed to move around, map the area, and know where to pick up and drop items.
- The Robotic Arm: This arm, with at least 5 degrees of freedom, needed to lift at least 2kg and place it precisely where required.
We started our journey using ROS (Robot Operating System) and quickly realized how frustrating and inefficient it was in handling various situations. (We might write our own OS, we have started writing our own modules wherever extremely required).
After designing and building a simulation using ROS, we searched for motor vendors. To our surprise, there were fewer than five motor vendors in India (we found only two) that produced BLDC motors for robotic arms or drones. It took over 45 days to narrow down our options, but the delivery times were between 20-25 days. We hadn't even gotten to motor controllers, drivers, and encoders. It was faster to get motors from the USA through Amazon than from a nearby state in India, though it was not cheap.
We faced many other reality checks, realizing that the supply chain and manufacturing ecosystem in India wasn't ready for us to immediately build the machine. While it was naive to believe we could accomplish this easily, it was also brave and ambitious to attempt building the product in India, trusting in the Make in India movement.
If you want to build in robotics, the USA and China are the best places to be.
Hopefully, things will change in India in the coming decade 🤞
The upside of this journey was that we had the time to thoroughly explore the market and various categories with a fresh perspective, leading to amazing new ideas and opportunities.
Speaking to our mentor reminded us that we are on this journey for the next 10 years, and we need to start building our bridge to reach our desired destination.
We have identified two key problem statements that we are currently working on, and we will be releasing demos and providing early access to gather feedback and find stakeholders to collaborate with.
This approach will help us gain necessary market validation, become more integrated into the robotics community, and attract early customers to work with.
We don't want to build in isolation or believe we are doing miraculous work without real-world feedback. If an idea doesn't have customers, you can't build a business around it. Period.
PS:
I delayed writing this blog because I wanted to share an exciting demo video, which would provide the much-needed dopamine kick. However, building in hardware is challenging, and the problems we are trying to solve will take time, even if you work passionately for 90-100 hours a week.
I am having the best time of my life working with Sanjana. Every day, I wake up with the hope of doing some really great work. She inspires me to push through tough days, and I pull her out of the rabbit hole when she goes too deep & it's unnecessary.
We will reveal the problem statements with some demos soon, aiming to release them for early access in a couple of months.
If you are contemplating starting something, consider robotics. You won't regret it. There are numerous problems to solve, and the potential impact is enormous.
PPS: At the halfway mark of 2024, one of my goal for the year was to build some world-class products with active users. I am confident we will be releasing 3 products by the end of this year.
Subscribe to this newsletter for more such blogs. Follow me on Twitter for the BTS of building.