Hardware is seeing a resurgence and exponential innovation similar to the software boom a decade ago. And as a robotic delivery company founded in 2017, we have been navigating the manufacturing world and hardware revolution, picking up product design and smart tips along the way. Applying new technologies to improve the manufacturing process guarantees quality, profitability, and lower cost. However, many Internet of Things (IoT) devices and systems still lack automation when it comes to the manufacturing process, with human intervention still being key. And IoT startups often don’t realize how fast they will grow and scale – this is when close relationships with a network of specialist manufacturers become a necessity. Here’s an overview of our learnings. Human Touch and the Industrial Internet of Things (IIOT) Co-Existing By far the most obvious candidate for automation is the manufacturing sector due to the repetitive, predictable tasks, from packaging products to loading, and the sheer demand that came from the e-commerce boom. However, manufacturing isn’t the top contender; rather unsurprisingly, it is the food delivery industry, the sector where our semi-autonomous robots are deployed. The robots have their own software and are made with carbon fiber cases, silicon boards for electrics, LED screens, high-tech light sensors, and cameras to improve autonomous decision-making, path planning, and obstacle avoidance. However, our manufacturing process – which could employ a network of sensors to collect production data and use cloud software for valuable insights about manufacturing efficiency – still lacks some automation, which is common across the IoT industry. With robotics, everything is made with highly sensitive silicon, high-tech, and hardware that needs close human intervention to review prototypes. This learning curve is so that our process will never be converted to lights-out manufacturing – and other IoT companies should not feel the pressure to do so either. unpredictable The most critical assembly stage for the robot is debugging, which still has to be done manually, and this is why it takes four hours to build. At Kiwibot, the mid-term vision is to work to half the manufacturing time by automating the debugging stage and focusing on pre-assembly. Here, computer-integrated manufacturing and AI technology would be key – we are planning to connect our manufacturing partners to our AI systems to improve quality assurance processes and reduce human error. It is essential to be ahead of the curve since the IIoT was from $12.67 billion in 2017 to $45.30 billion by 2022. projected to grow Efficient Robotic Manufacturing: Modular Design and a Network of Partners Last year, we designed, prototyped, and manufactured our next-generation robot in less than seven months during the pandemic, with team members across Taipei, Medellin, and Shenzhen, China, with thirteen hours of time difference. Internal communication was only possible through asynchronous tools, such as project management software and process documentation – and, most importantly, a shift to . This allows iterations and updates to robotic designs and manufacturing plans without affecting the production flow. Modular design also enabled us to move forward with the more complex parts of the robot, even when the simpler features weren’t yet figured out. modular design The pandemic also meant we could not visit China for manufacturing tours to perform quality assurance processes or review different product iterations. This led to various challenges with external manufacturers, vendors, and partners. Once, Nvidia AGX, the brain for the robot, was out of stock while we still had 400 robots to build. The key trick was to leverage a network of manufacturers and not just rely on one. Kiwibot doesn’t need a lot of parts, so we can’t work with big manufacturers due to quantity and the particular sensors and boards we use. This is one of the common pain points for businesses as many are small-scale and can’t afford the high prices of manufacturers. Therefore, to source new parts, we turned to Alibaba as a hub for vendor information. It now offers SMEs a “digitalized end-to-end manufacturing supply chain that allows for fully-customized, .” demand-driven production As we deploy delivery robots all over the U.S., we can now have conversations with larger manufacturers, with their HQs mainly in Colombia. We aspire to be a leading designer of robots and improve the quality of our delivery services. But, we do not intend to be a leading manufacturer. Our goal is to find best-in-class manufacturing partners that can materialize design proposals. Building Smarter Cities By teaming up with manufacturers who use the IIOT system and to evolve the production process, we’ll be able to analyze the data to improve the next round of productions and create devices to assist the growth of innovative, green cities. Designing smart robots through partnering with smart manufacturers means better, low-cost products, and in turn, lower costs for our customers, whether local authorities or small businesses. compile data through AI Since it is predicted some of all manufacturing activities could be automated, people will be forced to look for jobs that are deemed higher value and needing human intelligence. Those jobs will be better paid because the manufacturers already reduced the cost of low-wage jobs using automation. Take the shipping department of a manufacturing company, for example. So many resources are now spent on delivery and transportation. But if the prices were reduced, people’s share of wallet (SOW) would be smaller, so they would order more. 59 percent When cars were introduced in the early 1900s, people protested to protect the horse and cart. The same argument was used then; that the car was going to take away thousands of jobs. But just look at the since the car invention – the prevalence of motorized vehicles, or in our case autonomous sidewalk devices, are often viewed as a visual metaphor for development. progress recorded (GDP per capita) Production is only the first step to smarter cities; getting that product to its destination safely and intact can be a make-or-break moment for a manufacturing business. As we scale our production of robots, we plan on developing packaging that is resistant and aligned with our key robot design points. These include shipping from our partner’s factory to the U.S. and shipping from the arrival warehouse to our client’s locations like campuses to roll out our latest innovations. As the world continues on the digital transformation journey, new manufacturing, supply chains, and automation innovations are popping up. As autonomous vehicles and sidewalk delivery devices are set to be the most disruptive technologies of our time, it only makes sense that the manufacturing behind them follows that lead.

Flow

Chain

Assembly

NVIDIA

2022 - HackerNoon Contributor of the Year - Robotics

Nominated for 2022 - HackerNoon Contributor of the Year - Robotics

Too Long; Didn't Read

Developers: Build Less. Deliver More. [Learn how]	

Building Robots for Smarter Cities: How the Manufacturing Process is Everything

Too Long; Didn't Read

Companies Mentioned

Coins Mentioned

Natalia Pinilla

About Author

TOPICS

Languages

THIS ARTICLE WAS FEATURED IN...

Building Robots for Smarter Cities: How the Manufacturing Process is Everything

Too Long; Didn't Read

Companies Mentioned

Coins Mentioned

Natalia Pinilla

About Author

TOPICS

Languages

THIS ARTICLE WAS FEATURED IN...

RELATED STORIES