The Next Big Thing
Posts about next generation technologies and their effect on business.

Automation prioritization

 

Since I have been putting out some posts on automation, I’ve been getting some feedback from coworkers. One was a post titled: 8 Questions to Ask before You Automate. It holds some useful perspectives to evalutate if automation is even applicable to a situation.

 

Back in the early 90s, I led a project called Knowledge-based Tool Design focused on improving the productivity of tooling designers for automotive manufacturing. We used the capabilities of CAD tools to try and automate as much as we could, related to the creation of the machines that facilitate car assembly. This was a high value effort, since late stage engineering changes in the car parts themselves have implications on the tooling and can delay the startup of manufacturing. Anything we could do to address the reengineering of tooling had a direct effect on time-to-market.

 

We would load the car component models that need to be clamped and welded into the CAD system and try to create automated techniques to define and design the tooling needed. After firing rays all over the place to determine entry routes for robotic arms holding the clamping and welding tools, it became clear that people can look at parts and tooling and determine routes for entry very easily compared to doing this programmatically. Creating these designs well definitely involved creativity and intuition.

 

What people could not do reliably was define the underlying Bill-of-Materials need to create that robotic assembly, physically. So yes, I learned back then that it is very useful to understand what people and/or computers are good at, when defining the right approach to address repeatable, higher-value, computationally capable tasks with automation first.

 

Whether it is designing tools, answering calls or writing software – even though the automation capabilities are radically improving, this assessment is still required. I usually think of it as a 3 dimensional matrix and the further away from the origin, the more likely the automation effort will be effective.

 

 automation axis.png

Just because we can do something, doesn’t mean we should do something, especially when there is a constraint on the effort available to tackle a set of tasks. We need to prioritize.

 

 

What is automation?

IT automation.pngLately I’ve been talking with teams about ‘automation’. I am beginning to think that automation activities are a lot like innovation for two reasons:

  1. The definition is very dependent on the role of the individual
  2. It can be identified by: I’ll know it when I see it

My definition is:

Anytime a computer/system can do something where a person adds little to no value

 

It’s not just robotics or cognitive computing, it can be as simple as automatically updating information between systems (e.g., master data management) -- so people don’t need to swivel between systems to perform updates.

 

IT related automation can range across a variety of areas:

  • Plan
  • Define
  • Build
  • Run

The role of people and their expertise is critical and one of the things we need to determine is where their attention would be better served and how automation can enable us to maximize the value from it. Each of these areas has a range of tools that are related to incremental or disruptive improvement.

 

The computational capabilities are there to facilitate significantly greater automation, we just need to understand our options. Strategically the options and impact are going to be immense. Cloud computing is essentially enabled by the automation of infrastructure deployment and management. We now need to take it deeper into the Application Portfolio and the business processes themselves.

 

Will there be a new dimension to UI design based on cognitive computing?

things.jpgI recently read a report on How Humans Respond to Robots that focused on the social side of robotics. Near the end of the article it started to talk about autonomous cars, autopilots and other devices that are really robots but we don’t normally think of them that way (e.g., smart thermostats) because of their minimal user interface.

 

In a world where we’ll soon have wearable technologies all around us, likely building to a $50 billion market by 2017, pumping out data to feed the autonomous response of cognitive computing, it does make me wonder how enterprise architectures and application portfolios will drive us into the uncanny valley for business automation.

 

As we automate things and find patterns, the automated response could be a bit eerie and unsettling. Likely a new dimension for User Interface Designers of the future will look beyond just the traditional capabilities and more deeply into user intent and the sense of user interactions – so we can accept the assistance as intended.

Is it time for a Chief Automation Officer?

Automation officer.pngOver the last few years, there has been quite a bit of discussion about the race against the machines (or the race with the machines), based on the abundance of computing available. When I think about the IoT and its implications on business, it may be that information is just a side effect of an entirely different corporate strategic effort.

 

Maybe there is a need for a Chief Automation Officer more than a Chief Information Officer going forward?!? Someone who looks at the business implications and opportunities for cognitive computing, sensing, robotics and other automation techniques.

 

Or is automation just assumed to be part of all future strategic planning activities. As I began thinking about it, it’s clear that others have thought about this CAO role as well, although mostly from an IT perspective instead of one based on business need. It could be viewed that this is a role for the CTO or even the enterprise architect.

Context, automation and the future of services

looking for direction.jpgThere recently was a story about a computer program that passed the Turing Test. When you get into the details of what was actually done, I am not sure it really qualifies. The fact that people are talking about the event though is enough to show that we’re pretty far down the road toward breaking down the perceived barriers between machines and human interaction.

 

These advanced levels of interaction capability are enabled by a new wave of AI applications that can capture context at scale and in near real-time. These solutions when they move out of the labs should be able to consume massive amounts of information and generate contextual understanding at a level that even the most intuitive individual would find difficult to match.

 

You might ask what does this mean for the future of services. Or where will it be of use to my organization? It should be applicable at just about any point where a conversation occurs with customer or between:

  • employee and employee
  • organization and organization
  • government and citizen

We may be able to automate interaction that isn’t face-to-face and even then it may need to be person to person with the likelihood we can overcome the uncanny valley.

 

These new context-aware, AI enabled interactions can provide a multi-level view on engagements and ‘experience’, allowing organizations to filter through the noise and latency (for example waiting for certain skills -- Spanish language) and shift the focus to an enriching experience, relationships, and achieving goals. I can easily see a future talking with an AI agent at the drive-up window, as a low-hanging opportunity.

 

The recent book The Second Machine Age, examines how society, the economy, and business will transform as digital technologies and smarter machines increasingly take over human occupations.

 

It makes you look for direction about who will robots put out of work? This interactive graphic from Quartz takes a stab at answering that question—exploring which U.S. jobs are most likely to become automated, and how many workers could be affected.

Search
Showing results for 
Search instead for 
Do you mean 
Follow Us
Featured
About the Author(s)
  • Steve Simske is an HP Fellow and Director in the Printing and Content Delivery Lab in Hewlett-Packard Labs, and is the Director and Chief Technologist for the HP Labs Security Printing and Imaging program.
Labels
The opinions expressed above are the personal opinions of the authors, not of HP. By using this site, you accept the Terms of Use and Rules of Participation.