How to get 20 times more productive – ask a robot

Human beings push the boundaries. They can achieve the most astonishing feats. Take Lucas Etter, a fourteen year old from Kentucky USA who, in November 2015, created a new world record by solving a Rubik’s Cube in 4.9 seconds. An amazing achievement which you or I would have no hope of repeating. Even Lucas would struggle to reproduce his achievement regularly though he was able to explain how he had broken the record – by memorising an algorithm that defined what he had to do and by training his spatial skills and muscle memory. Lucas had turned himself into a ‘well-oiled machine’ for solving the Rubik’s Cube problem.

Just three months later, Jay Flatland and Paul Rose revealed their approach to the same problem. Their solution reduced the record to 0.9 seconds – over five times faster than Lucas.

It wasn’t really a fair contest. Jay and Paul had built and programmed a robot that replicated Lucas’s approach. Built from fairly standard components, the home-built robot used 3-D printed parts, stepper motors, and an Arduino chip, which was connected to a Linux system. The algorithm recognised the starting state of the cube and implemented a series of stepper motor movements to reach the solution as quickly as possible. And, of course, the robot could repeat the task again and again, without any degradation in performance, in contrast to Lucas, whose brain and attention would quickly wander and whose hands would inevitably get sore.

20 Times Increase in Productivity

Even if Lucas could consistently repeat his world-record performance, he would still only be limited to around 29,000 Rubik’s cubes in a 40-hour working week. On the other hand, the robot who is not limited to ‘working hours’ can solve one every single second. This would result in over 600,000 completed Rubik’s cubes completed by the robot – 20 times better than Lucas. If solving Rubik’s cubes was your business and your generated £1 for every one solved, this would increase your revenue from £29k to £600k per week. Now, wouldn’t that be transformational?

Screenshot_08_03_2016__17_06

This is just one example of a situation in which a real machine is able to beat the most expert human. Many people are realising that such situations are found frequently in businesses and companies throughout the world. They are questioning the balance between human workers and specially programmed robots. Even the most expert employee may not be as productive as a robot. Their performance may vary, they may tire and quit. In contrast, a robot continues to work indefinitely and with minimal cost. Every business owner knows the complications of replacing a worker, especially if that worker is the best in the trade! If your robot breaks, a new one can be installed quite simply.

Predictable and Repeatable

In the Rubik’s Cube example, with a few low cost components and less than 40 lines of Kociemba algorithmic code, Flatland and Rose managed to create a robot that worked 5 times faster and produced 20 times more output than the global expert. Is the same approach relevant more widely? Is it possible to apply technology to a range of jobs or services to increase productivity tremendously? How widely can algorithms be deployed?

An algorithm involves a set of rules that ensure a predictable and repeatable outcome each time it is called on. Not too long ago, algorithms and robots were felt to be limited to very repetitive situations. However, in today’s world, it is not just predictable and repeatable tasks that can be automated. Autonomous vehicles are one such example, and we are well on the way to ‘robodoc’ (robots that assist with surgical procedures) and exam marking by robots.

The principles of automation are not restricted to hardware and robots – they are equally relevant to software. Perhaps the most common example is the email sent from an online shop each time you make a purchase on the internet. Obviously, the confirmation email is not written by a person, but is generated by a computer using a program that gathers a few key pieces of information (such as product details, credit card details, delivery preferences, and etc.), and constructs an appropriate email using a defined algorithm. The cost is tiny. Just imagine how many humans it would take a large retailer to do the same job. And imagine the costs of the office staff involved!

As globalisation continues, the need to stay ahead of international competitors becomes ever more important. It is essential to increase productivity whilst maintaining quality and, given high labour costs, this can only be done through automation. Take a moment to reflect on all the tasks carried out in your business each day. Which could be done better and cheaper using automation? You might be able to reduce costs, while reducing response times, and increasing productivity and service. If it works for cars, for delivery and invoices, and for Rubik’s Cube, it might work for you.

Photo credit Patrizio Cuscito.

Catherine Steele

Over the past 15 years, Catherine has grown the business to become a respected and trusted global voice for business-led technology consultancy. Spaarks operates in the UK and Australia serving an impressive client base. Prior to forming Spaarks, Catherine was responsible for Equities Research IT systems at JP Morgan-Cazenove and held positions with Aegon Asset Management, The Open University, Sopra Steria and Wood Mackenzie. She has a Masters Degree in IT, an Undergraduate Degree in Philosophy and now lives near Aberdeen with her husband and twin children.

Leave a Reply

Your email address will not be published. Required fields are marked *