Risks of automating smart decision-making

When working manually, not enough time is given to carefully consider and find the optimal solution to the problem. automation, in turn, is able to develop a greater number of ways to solve the problem, analyze each of them and choose the most optimal for each specific situation.

Among other things, automation helps to assess the state of debt and develop preventive measures to improve it, as well as quickly plan and manage payments to the organization, daily analyzing its liquidity and optimization.

The risks of automated decision-making are poorly understood, but we can clearly identify the following factors that can cause serious additional difficulties in solving the problem.

Like any complex system, smart is prone to breakdowns. At the same time, due to the fact that the idea of automating decision-making is still developing, today there are not many companies that can accurately assemble the system, while meeting the deadline and the stated cost.

In addition, there is a problem of hardware incompatibility, which can only be avoided by a specialist who understands both the modern features of automation and the software that is required for it. If in the West more often sell a ready-made system, in Russia it is assembled by separate components, which creates an additional risk of errors in the configuration and Assembly of equipment.

There is also the "Domino effect" problem. There is a high probability that if only one device fails, the overall chain and operation of the smart system as a whole will be disrupted.

An important aspect that is inevitable in the era of digitalization is information security. Every year, technology becomes more complex, and hackers become more sophisticated and skilled. Many "smart solutions" are vulnerable — they can be hacked, intercepted, which can lead to the most unpredictable consequences.

These risks require special attention when making automated "smart" decisions. it is necessary to improve methods of countering cyber attacks, constantly update the system and conduct its prevention for breakdowns and failures.

In case of introduction of automation of making "smart" decisions within the company, it is necessary to check its readiness for this process, to make sure that it meets the following necessary criteria:

First, the company's project goals should be clearly defined and measurable, and priorities should be clearly set for achieving them.

Secondly, the project requires separate budget funds, as well as an understanding of where they should go: for the purchase of licensed software and equipment, payment for consulting services, for specialized training of employees, etc.

Third, to successfully launch automation of smart solutions, you need a project team that will fully understand and share the project goals and priorities. It should consist of specialists whose professional level corresponds to the goals of the project being implemented.

Fourth, in order to avoid the risk of sabotage of the company's managers, its key employees must be motivated to achieve results.

The risk of inconsistency between the company's strategy and the automation strategy implemented in it is in principle the most obvious. In this case, the company will quickly face the difficulties of their inconsistency, limitations of existing equipment and the need for extraordinary investments.

In General, we can say that the set of errors made when introducing automation of decision-making in the enterprise is quite standard: these are problems of goal setting, implementation stages, and project management.

Such "typical" risks as delays due to perfectionism and a desire to bring the project to "perfect", dismissals due to innovations, risks of staff not being ready for changes, lack of responsibility and professionalism of project teams do not disappear.

This can also include a superficial attitude to entering initial data into a "smart" automated system, saving on the number of automated jobs and equipment, underestimating the time and cost of implementing technologies, and other obstacles for which it is necessary to think through solutions in advance.

The better the automation implementation process is set up and calculated, the higher the probability of its success and the expected course of development.

Source: https://sci-article.ru/stat.php?i=1577471894

Posted: September 2020