How Do You Define a Robot? - Blue Prism
+44 (0) 870 879 3000 | +1 888 7577 476 (1-888-75-PRISM)

How Do You Define a Robot?


How Do You Define a Robot?

In this month’s contribution to WIRED’s Innovation Insights, Blue Prism Chairman Jason Kingdon shares his definition of a robot. Jason explains that a “robot” is anything that can mimic human tasks – regardless of its form (hardware and software).

To read more about Jason’s take on what constitutes a robot, click here or read below.


What Defines a Robot?

The EU Robotics Assembly has now agreed to disagree on the definition of a robot. The complexity of this question was not lost on a group of 10-year olds attending Imperial College London Robot Summer Camp.

The junior experts explained there was a range of definitions. While some insisted a robot must be able to move (and so discounting a washing machine), others felt it was the machine completion of a task that was key. One thing that both groups did agree on is that there must be a capability for re-training.

In terms of software robots, the EU Robotics Assembly has a point – a robot is somewhat different depending on context. An automation script is a robotic completion of a task. It is re-programmable (if not technically re-trainable), but lacks impact in a crucial dimension that seems important in this series – the ability to carry out multiple tasks above, beyond and surrounding the core automation.

What seems exciting about the new wave of robotic software is the concentration on mimicking the human agent. The advantages keep racking up in adopting this approach. It means that forward and backward compatibility is built in to the core robot design. If a human can use a software tool, then so can a robot. Of course, if a software application is also designed to be used by other machines trough an API, it is doubly useable by a robot! The freedom and efficiency this offers from a software standpoint cannot be stressed enough.

Equally crucial is the robot’s starts to make gains on a major human advantage – flexibility. Humans do such diverse things. Confining ourselves to humdrum office admin tasks, humans effortlessly link between an HR system and a compliance process; or an accounting system and expense log; an online banking to bespoke company budget system; a marketing system to the CRM. These simple steps have historically been a major IT challenge. Software robots change this.

This ability to mimic the flexibility of humans seems a core advantage, as does what they do over and above carrying out of the main task. The 24 hour-a-day, 365-days-a-year work rate is a given. As we have been discussing, the less obvious are all the collateral benefits the robot brings. For example, scalability: need more throughputs? Add more robots: no hiring, no training, just more. Think of those planning sessions saved.

Documentation: we all live and breathe it; all commercial organizations must do their share. We’ve heard about “frictionless selling” – what about frictionless administration? You want to understand what a process is doing? Get the robots to churn out the process flow charts, audit logs and security updates across the whole process diagram.

What about frictionless analytics? It’s all very well to see the process flow charts and understand how a process operates, but what really takes place? Robot: Show the management information across the live process flow. Live? Surely. A historic sample of the process flow? No, a software robot can show the live event, in situ. Live flow charts highlighting the individual bottlenecks and workflows across existing core IT. Frictionless analytics.

All of which can then be used to iterate better process designs. This is 21st century operations management. All of these benefits are basically “free” built in as standard, which contrast to where we have been where all of these areas require dedicated kits. Robots seem to be most ruthless in terms of the extra jobs they do; hording and expanding their uses.

An example from the physical world highlights this. Baxter robots are mechanical robots (unlike our clank-less software versions) and assembly line specialists, or at least that was their genesis. The robots are workforce multipliers, trained, not programmed, and have extensive capabilities around manufacturing. However, Baxter robots are now being used as surgeons (for doctors to carry out remote operations), robo-farmers minding crops, in space, and now as paraplegic assistants. That is, a robot-enabled wheel chair.

So, something imagined for industrial assembly is ending-up as remote surgical tools, astronaut farmers, and now an aid-worker for human mobility on Earth. The chair, requested by David Whelan, who was left quadriplegic after a skiing accident, uses an adapted music control to allow instruction to the robot; so getting things from shelves, picking up dropped items, manipulating other human designed objects (maybe the vacuum cleaner?) as part of everyday living.

This is another version of the human interface issue, and exactly the kind of imaginative dynamic unforeseen benefit that flows from “great” general purpose technology. How far and how quickly this all goes is more constrained by frictional imagination and ingenuity.

The new robotic chair allows a person to collaborate with a robot forming a human-robot partnership. A case of truly racing with the machine.