Automation everywhere - a look at what UK organisations are automating and why
In a technological society automation is a necessity, but care must be taken in its rollout
Automation, the continuous process of removing human intervention from the delivery of products and services, is a cornerstone of technological progress, a driver of the economy and the focus of many organisations' digitisation, productivity and security programmes.
It's also something that can produce anxiety in those who fear being displaced by machines. Overall, technology has been a creator rather than a destroyer of jobs thus far, but this big-picture positivity will not be shared by those being shunted aside, who may not have an equivalent post to move into.
In many cases, though, automation is a necessity rather than an option. Time moves on and organisations must move with it, but working out what to automate, how and when can be a delicate balancing act.
What are organisations automating?
A recent Computing Delta survey of 120 UK IT leaders found that cyber security and incident management was by some distance the most prominent focus of automation efforts in organisations of all sizes and sectors, no doubt in part because of the move to remote working.
What is the main focus of your automation efforts?
Base: 120 UK IT leaders
Next was the introduction of robotic process automation (RPA) to handle some back office processes and customer support operations. Then came business process automation (BPA), automating production systems, software deployment and development, and employee and customer self-service projects. Of course, many interventions will be a combination of these topics.
In the SOC
Cyber security is one area where automation has long been a necessity rather than an option. The volume, intensity and speed of attacks are increasing all the time, and threat actors are frequently automating their pocesses to hit as many targets as possible, inflict more damage or avoid discovery.
A programme manager in the public sector even said "the aim is for zero staff intervention with all security managed by AI".
Whether we'll ever be able to turn protection against threats completely over to AI is moot. Most IT leaders talk of AI augmenting human capabilities rather than replacing them - the exception being lower-level operatives doing manual and repetitive tasks who might do well to update their CVs.
Far from being intelligent and all-seeing, AI can be spectacularly stupid if applied to the wrong tasks or trained on poor quality or biased data. Where AI/ML is currently successful, it's still mostly in specific, narrow, data-intensive use cases such as scanning for vulnerabilities, alerting and segmenting threats, and base-lining network operations.
A catch-all automated defensive system requiring zero staff intervention would seem to be some way off. However, AI is advancing apace, and future change will likely not follow a straight a linear trajectory but rather a hockey stick curve where progress suddenly accelerates, at least in particular domains. So the unstaffed SOC may not be as far off as it seems, and security analysts may ultimately have to seek pastures new.
In the back office
Some back office workers too might be looking over their shoulders.
"We're looking to reduce back office headcount via the use of RPA on transactional activity," said a pro vice chancellor in the education sector.
This person may simply be stating what's often left unsaid: there is a clear economic incentive and potential efficiency gains to turn repetitive jobs over to machines. The more usual story, though, is that RPA augments back office work by reducing the drudgery of paperwork and form filling and thus increasing productivity. Peoples' jobs will change, but not be lost, with latent talent unleashed and made available for higher value tasks.
RPA was a popular option for other respondents too, including in the delivery of self-service, both for customers and for staff. The benefits of this are obvious, particularly in the age of working from home, where human service is at a further remove.
"We're undertaking digital transformation initiatives to enable online customer self-service and remove manual back office functions," said an IT manager in a mid-sized manufacturing company.
"We're using tools like Microsoft's Power platform to automate processes for office staff" commented an IT director in the public sector.
Meanwhile, others were rolling out low-code solutions enabling staff to automate their own workflows.
In the pipeline
Software development and deployment is another area where automating manual processes can make a big difference to productivity and, in this case, code quality and security. Automation is at the root of DevOps and the further you go into DevOps the more manual and repetitive tasks are handed over to scripts and machines.
"We're automating DevOps pipeline, security and testing," confirmed a director of agile and DevOps transformation in a bank.
Large-scale business process automation tasks are under way in many organisations too. "As part of a BPO arrangement and the successful migration of ERP to cloud, business processes are to be automated through robotics," said a head of infosec in the energy and water sector.
In the cloud
Unsurprisingly, many of the automation projects mentioned are being delivered through cloud-based services. 62% of respondents said procurement was via the Opex/SaaS route, with 38% saying it was primarily funded by capital expenditure.
Automating production systems and backup and recovery were more likely to be funded through Capex, while RPA and cyber security leaned heavily towards the SaaS/Opex model.
In terms of partners, 45% of respondents said they work directly with vendors, but more than half bring in one or more managed service providers or consultancies. Automation projects are often complex and outside expertise is much in demand.
Automation can be a tough thing to get right, not least on the human side. Some people fear it will be the end of their job, and in some cases they will be right. Historically, automation has tended to create jobs in the wider economy, but the job creation is lumpy and those displaced back office staff and technicians may find themselves without equivalent roles to step into. So those very human attributes of empathy and creativity are most definitely required when rolling out automated systems.
Done right, RPA, AI, natural language processing et al. really can take on those tasks that no-one likes, alleviating pressure, improving consistency and productivity. But organisations need to work out how to get the best out of those employees who are thus freed up to use more of their innate skills.
That is going to be one of the biggest challenges as machines take on more and more of what we used to do ourselves.
For in-depth discussion on automation and cloud join us in September for Deskflix: The Future of Cloud Automation. Register for free today.