Optimisation for robots, not for people replacement

The weak economic growth creates a growing driver for cost reduction and the demand for such solutions. Progress in AI and robot capabilities and growing offers of such products on the market generate a supply on the other side.

That combination looks perfect for the next wave of improved efficiency, increased productivity and division of labour.

What is required to convert that opportunity into positive results and success?

There are two approaches to automation:

1.       The first is a new Xiaomi “dark factory” https://www.youtube.com/watch?v=DVzhjkX7bVU&ab_channel=TechPlans.

2.       The second is one of Amazon's warehouses https://www.youtube.com/watch?v=ZWonAz7Kczs&ab_channel=NBCNews.

The Xiaomi “dark factory” implements manufacturing and operation processes specifically designed and optimised for robots.

The Amazon automated warehouse replaces people in workflows, operational processes and an environment initially designed for humans.

What approach is better? A summary is below

Xiaomi	Amazon
The solution needs complex enterprise design, including an operation model, technology platform, and full-scale implementation and testing. Waterfall methodology.	The solution has a short time to market, when components are available, and can be deployed incrementally. Agile methodology
Complexity is balanced between the overall solution and robotic components.	Complexity is mostly in robotic components.
The solution is optimised for high productivity, high quality, low operation cost, and flexibility.	The solution is optimised to replace people in the workflow gradually.
The solution is stable after deployment.	The solution constantly changes the workflow and operational processes until robots replace humans.
The solution is optimised from the beginning.	The solution inherits sub-optimal processes, operations, topology, equipment and environment from the “human stage”.
	The solution is exposed to operational human risks and errors and requires complex management.

The main advantage of the Amazon solution is that it is SIMPLER to start.

There is no need for complex enterprise design, development and testing. As with any Agile project, risks are distributed across the project delivery timeline. Reporting is comfortable, and many milestones can be defined and achieved on time.

But that is about it.

All other parameters are more beneficial to the Xiaomi solution.

There are two additional key factors to consider.

·         The Xiaomi solution is a classic and well-known automation of manufacturing processes in the manufacturing industry, with its broad experience with industrial robots, from car assembly to robotic conveyors.

·         The Amazon solution is for a warehouse where automation is not mature. Still, the transport platforms on wheels, as shown in that video, look much more suitable and rational on a smooth floor than walking robots.

The conclusion for this case is:

The optimisation of operational processes for robots is more efficient than the optimisation of people replacement.

The Amazon warehouse is not that rigid to be optimised for industrial robots. Shelves, aisles, and lights also look excessive and suboptimal for robots but can be reorganised quickly and relatively simply. That optimisation can utilise the Nvidia technology of Digital Twins to reduce cost and shorten time to market.

The optimisation for robots would make the transition process more transparent, more honest, straightforward for employees, and more manageable for management.

Optimisation for robots does not need to be as expensive as the Xiaomi “dark factory”.

If key industry players and vendors define standards for robots and elementary operations, the operational process could be assembled as simple as Lego models but still have flexibility for individual solutions.