Business efficiency and optimality (in the name of which automation is being carried out) over time have become elusive goals that are constantly moving away as the scale of automation increases.
At the same time, the stage of digital transformation is qualitatively different from the previous stages of the introduction of new information technologies, digitalization, etc. The promotion of ideas and methods of digital transformation, under certain conditions, can become fruitful and radically change the way of doing business, as well as the methods of management used.
Justification of promising areas of work on digital transformation of business is correctly associated with fixing the accumulated problems: “evidence can be perceived only by those who have a common past experience” (NI Vavilov).
1. “As is” does not provide guidelines for “To be”.
An unbiased audit of business processes and the level of business automation will most likely establish the incorrectness of existing regulations (they are out of touch with reality) and the vulnerability of decision-making practice (without correct data and without taking into account the problems of adjacent departments). In addition, the unspoken general conclusion will be that managers do not quite know their business processes.
During the audit, many positions can be identified that seem unreasonable, controversial, incorrect, etc. However, it is difficult to implement the idea of ​​eliminating the identified problems, since the audit itself does not provide guidelines for business improvement and criteria for recognizing the correctness of the decisions made.
Only this business itself can understand how to change (develop and improve) a specific business and make the necessary decisions: this is its money and risks. External consultants can only bring new tools to the business. The current situation is complicated by the fact that the necessary tools do not yet exist that would allow a business to look at itself objectively, without following the stereotypes that have already formed in it.
2. External directive restrictions on business.
Government regulation imposes external policy constraints on business. This is neither bad nor good - it is. Sometimes business finds ways to compensate for their negative impact (offshore, etc.). But very often, such constraints become a burden just because the business does not generate countervailing compensating opportunities. Consider a simple example of the impact of a staffing requirement.
3. Organizational structure: division of labor and cooperation. Example.
The staffing table should be in the company for many reasons. For employees, the staffing table in the form of an organizational structure is important because it fixes their status in the company. For business, the organizational structure is important, as it records the division of labor and cooperation adopted in the company.
Moreover, the real division of labor and cooperation most often do not correspond to the nominal organizational structure.
The figure shows a graph of mail messages in a real company for 9 months. Massive internal mailings (notifications to more than 100 addresses) and unanswered external messages (like spam) have been removed from these messages.
Red dots correspond to external addresses, and blue dots correspond to addresses from company domains.
If we take into account regular communications (delete "one-time" letters) and stable mail flow paths, then the company's hierarchy, expressed through real communications, will be as follows (only mail from company domains):
Red dots correspond to more intense nodes, yellow - medium in intensity, blue - less intense.
Hierarchical subordination is manifested through the communication path (red line).
The company's communication graph, based on the relative autonomy of communications, is as follows:
From a practical point of view, it is interesting to reflect the current business processes in employee communications. The daily picture (random working day) of communications looks like this:
Separation of some communications is visible.
Reducing the time to 3 hours, we get the following picture:
For solving specific problems, a graph with the indication of mail addressees is more convenient. Red dots correspond to external mail, blue ones - mail from company domains. It is interesting that in this case, when solving a certain issue, the company's employees communicate with each other through participants who have external mail.
This example shows that the company has much more hidden knowledge than it supposes, and also that static regulations do not correspond much to real dynamics and practice. In addition, with the existing arsenal of tools, it seems very difficult to capture both "as is" and formulate wishes for "to be".
4. The software is a wolf in sheep's clothing.
In theory, the software makes it possible to organize a compulsory "rhythm order" of teamwork and significantly expand the functionality of specialists.
In practice, the software industry is a huge business with its own rules and economic model. The essence of the latter is to make universal (standard) programs and sell them to the largest possible number of clients.
Therefore, theorizing about how this or that solution will help a business, the vendor (manufacturer and supplier of software systems) simply sells a standard program (system) that clearly does not correspond to any particular business and which needs to be implemented. The implementation of the program consists in the fact that the code of the standard program is partially rewritten and the existing business processes and regulations are redone to comply with the standard designs (structural and functional) embedded in the program.
For the company, the existing practice of software implementation creates two latent problems.
The first is the possible loss of existing competitive advantages when there is no plan to create new competitive advantages (vendors do not understand anything about your business).
The second is the duplication of some business functions. Each vendor strives to reorient the client to their programs as much as possible, for which they include an excessive amount of functionality. This is especially true for accounting functions. Since the company does not manage with one program and one vendor, redundancy and overlapping of the functionality of the company's software landscape becomes inevitable.
A clear sign of trouble in the company's software is the volume of Excel files circulating in the company. In the most automated oil and gas industry, 70-80% of data is taken from Excel files when making decisions, rather than directly from the company's information system.
During the audit of business processes, it usually turns out that the programs used are doing something incorrectly or that data with an unknown status is used, i.e. "Non-systemic". In any case, most of the recommendations based on the audit results lead to the conclusion that it is necessary to expand the company's automation perimeter.
5. Summary of qualitative changes in companies as a result of automation.
The inconvenience of using existing programs leads to the fact that the entire business management system, to one degree or another, is displayed in Excel files "wandering" in the company.
The convenience of mail made it possible to formalize and record business communications. The imposition of automation has formed archives of historical and accounting data. The introduction of the business process methodology made it possible to present the business in the form of diagrams.
KPIs are widely used - metrics, it is not clear how they reflect the real state of affairs.
And finally, the passion for strategy initiated the formation of a unit responsible for creating the overall picture of the business.
The changes that have taken place in the last period can be usefully used, but so far, in general, they only generate costs. This is manifested in the fact that the bulk of management methods in companies, in fact, have not gone far from the practice of the second half of the twentieth century.
6. "Best Practices" - don't guarantee anything, but add determination.
BCG and McKinsey base their work on specially designed questionnaires, on the basis of which they prepare their recommendations. There are some differences in approach:
- BCG ;
- McKinsey .
It is important to understand that most of the consultants' recommendations are a generalization of specific proposals from the company's junior and intermediate employees. In fact, this also indicates that employees cannot directly convey their proposals for improving processes in the existing management system in an explicit and simple way.
To implement the recommendations of the consultants, special organizational forms are provided that replace the organizational management structures that have developed in the company for a short period.
The analysis of the state of the business by consultants inevitably reveals omissions and shortcomings that are advisable to eliminate. The problem is how, without having a general picture of the business (it cannot be obtained by surveys), correcting one does not harm the other.
The position of consultants on the formation of a business development strategy is very simple: this is not their business; they simply identified problems; they cannot make business development decisions for you; all they can do is introduce you to the “best practices” of similar companies in other countries and industries.
As a result, the company does not get a general picture of the business, or technology for regular business development.
Business is conducted in a competitive environment, therefore it is always accompanied by stresses and fears related to the vulnerability of the business, the possibility of detecting violations in conditions of conflicting legislation, violation of one's own status, etc.
Knowing that a competitor has "best practices" either breaches fear or becomes a new fear, but adds determination to change the business. As a result, “best practices” and new software are being introduced, which does not guarantee anything, but introduces the company into a new cycle of automation and implementation.
7. What's wrong with planning and management?
The main direction in revising the way of doing business in digital transformation is related to the revision of the relationship between planning and management processes.
The orientation towards the widely and long-propagated paradigm “long-term (strategy) - medium-term (year) - short-term (month) planning and operational management” has no prospects in modern conditions.
The idea of ​​drawing up long-term plans, their disaggregation and clarification cannot be realized with actual rapid changes in the situation and circumstances in which decisions are made.
At the present stage of business development, the majority of management decisions are made at the level of operational regulation. The main reason for this is the constantly changing environment, which cannot be foreseen even in short-term planning. This circumstance turns operational management into the main tool for increasing the efficiency of asset use.
The process of making operational decisions occurs almost simultaneously with the process of implementing business functions (production, repair, transportation), and the redistribution of resources and rescheduling are carried out daily, sometimes several times a day.
In these conditions, it is possible to "turn over" the existing hierarchy in the planning structure, taking operational management as the backbone.
At the same time, invariants are distinguished in the processes of the latter: digital assets, stable dependences and ratios.
For invariants on wider time horizons, structural and optimization decisions are made, which becomes an analogue of medium-term and short-term planning. Long-term planning at a new stage of business development is being transformed into qualitatively new approaches to take into account many factors and options when considering a development strategy.
8. The general picture of the business. Strategy. Operating business model.
The existence of a huge number of reports in a company does not mean that there is an overall picture of the business.
Strategy is a way of positioning a company in the asset space (existing and investment), in time and in the market.
The operating model of a business is associated with the creation of a "principle" business scheme in the context of the existence of numerous business processes in specific areas.
As a rule, the strategy and the operating model are static by the method of formation and are not associated with the constant analysis of historical and actual data (BigData). Therefore, they are not very useful for doing business, even if they are reviewed regularly.
And it’s not just a lack of good tools. The main problem is related to the fact that there is no unified system of production and commercial restrictions and management accounting: accounting ERP - yes, clients are somehow managed through CRM, there are GEO-business systems, repair planning and others.
If there was a unified system of production and commercial constraints and management accounting, then the strategy would be a dynamic vision of overcoming strategic constraints, and the operating model would be a way to ensure operational efficiency (situational constraints).
The intuitive desire to develop the areas of "digital assets" and "digital twins" is precisely related to the need for a business to have a big picture.
9. Purpose and method of digital transformation.
The concept of digital transformation has many interpretations according to what they want to sell you.
Moreover, it speaks for itself. A number is the most abstract concept introduced by a person. The top of the abstraction is zero, which means nothing. Transformation means transforming one thing into another. Not a replacement, but a transformation.
Digital transformation methodology is all about transforming what you have into what you need. Therefore, there is no need to replace existing programs with “progressively new” ones. You just need to have a way to renovate programs by using them in a qualitatively new way.
This way is the creation of a metasystem - a system over existing systems that does not change them, but brings its own new qualities.
The next question is the question of formalizing the concepts (objects) that business operates. These objects are very abstract: transaction, flight, purchase, repair, client, asset, investment, etc.
In contrast to abstract objects, existing information systems contain traces of mainly material objects: contract, invoice, route, spare parts, machine, book value, etc.
Therefore, it is necessary to be able to compare material objects and facts with the concepts and categories that business operates.
Oddly enough, but the backbone goal of digital transformation can be considered the creation of the SSDL (Special Symbolic Domain Language) language for describing goals, objects, operations (actions) and business problems. Forming a metasystem is only an intermediate means for creating an SSDL. In general, a specific business should have its own SSDL, taking into account the specifics and competitive advantages of the business.
The fact is that any system in constantly changing circumstances and requirements is temporary. Therefore, it is difficult to answer the question whether the system is good or not: yesterday it is good, tomorrow it is not.
With the language for a specific business, everything is easier: you can intuitively understand whether it is effective or not. If the language allows describing problems for which there is no understanding of the solution yet, then it becomes “eternal” (at least until the business dies out).
Essentially, all innovations from the stage of introducing new information technologies led one way or another to the creation of a language for describing a business: concepts, categories, objects and operations (actions). This manifests itself in the process of formalizing business processes, when agreeing on technical specifications and requirements for functionality, in determining the settings on which questions the reports should answer.
But as a result of long work, only a certain slang, or "non-language", of the business is usually formed. Sometimes, even worse, the business takes a "programming" point of view and begins to adapt to the capabilities of specific programs.
The history of progress unambiguously connects the creation of a specialized and convenient language (mathematics, chemistry, electrical circuits, genetics) with the beginning of qualitative development: it allows one to clearly formulate intentions, accumulate and transfer experience, and also unambiguously fix problems.
10. Digital asset.
In what follows, the term “digital asset” will be used in a broad sense. In a narrow sense, a "digital asset" is associated with business opportunities with cryptocurrencies.
A “digital asset” in a broad sense is what creates value in business at a new stage in the sharing of information and mathematized technologies.
Methodologically, the concept of a digital asset is intermediate both for the metasystem being formed and for the business language being created. It allows you to grope the path of digital business transformation. And if the attempt is successful, then a particular digital asset becomes both a component of the metasystem and SSDL (business language).
11. Production characteristics of the semi-finished product plant and three finished product plants. Example.
Plants for the production of semi-finished and finished goods are assets. Assets have production characteristics that are taken into account when planning the volume of production. Most likely, the downtime in scheduled repairs will be subtracted from the planned productivity of the plant and the expected volume of production will be obtained.
In dynamics, the volume of manufactured products, as a rule, will not coincide with the planned ones due to: actual terms of repairs, unplanned stops, etc.
The figure shows semi-annual data on the production of a semi-finished product (green 1) and the needs for it of three finished product plants (brown 2, orange 3, yellow 4).
The first figure corresponds to the data given on the same date, and the second - to the needs for a semi-finished product, recalculated taking into account the required time for its transportation.
Before April, actual data are given, and after - planned ones (too optimistic given the existing work experience).
The blue dotted line is the difference between the semi-finished product produced and the current demand for it.
The main problem is that the total production of a semi-finished product is significantly higher than the total demand for it. However, in fact, the semi-finished product is either constantly in short supply (purchases from third parties are needed), or there is an excess of it (it must be sold on the market). In the first case, there may be an increase in the cost of the final product, in the second case, it is required to deliberately inform traders about the volumes and the selling price.
Any of the graphs in the figure displays a specific time series. In a more familiar form, a time series is represented by two Excel columns, the first of which is a date, the second is the value of a certain indicator for that date.
What is a digital asset in this example? These are four (by the number of factories) different constructions, consisting of a set of time series according to production indicators (indicator - time - value) and circumstances corresponding to a specific time series (context of the situation).
In this form (this is not an exhaustive composition), a digital asset allows solving a number of urgent tasks. For example, based on the time series of receipt of a semi-finished product and the need for it, it is possible to determine (simulation model) the optimal size of the reserve, which minimizes purchases of a semi-finished product from third parties in case of supply interruptions.
In technical terms, a digital asset can be associated with a virtual object, Entity or BusinessObjects (in the form in which it existed before its purchase by SAP), that is, the innovations introduced take place primarily in the formalization of the representations of categories and business concepts, and not their software implementation ...
In general, a digital asset is a complex structure (object). A digital asset includes: historical data; economic and financial indicators; contracts and obligations; production plans and facts; peculiarities of work in different circumstances; analytical patterns (formulas); BigData from various sources (internal and external); mathematical models, including simulation and statistical.
The attack associated with the widespread use of Excel files outside the perimeter of the company's information system can be an advantage this time if the company's specialists in Excel files were able to express the essence of the functional elements of the business, avoiding unnecessary restrictions and assumptions imposed by typical software systems.
In most cases, it is quite simple (using a certain methodology) to form a first approximation of the company's digital assets based on "working" Excel files.
12. Management accounting and efficiency.
Methodologically, a digital asset is both a result and a technology.
As a technology, a digital asset corresponds to the way management accounting is formed: highlighting interesting data in different sources and processes, collecting it in one place, and dynamically updating it.
As a result, the digital asset contains the necessary data in the context of the circumstances, allowing to correctly calculate the efficiency: to isolate the costs of interest and divide the reasonable total costs when calculating the efficiency.
13. Digital twin. Model of the GOK concentrating plant operation. Example.
An enrichment plant of a mining and processing enterprise (GOK) can be represented as a system of series and parallel connected modules (nodes) with varying characteristics of throughput and indicator values ​​depending on the operating mode.
Scheme of the concentrating plant GOK.
The input data comes in different positions and corresponds to historical or projected time series.
The following are the simulation results for various circumstances.
The results consistently correspond to the stages (red number on the graphs):
- crushing and transfer unit;
- main conveyor;
- medium crushing;
- screens;
- separation of ore after screening into two streams: into small crushers and onto a conveyor for a blending warehouse;
- small crushers;
- conveyor to the blending warehouse;
- separation of ore into two streams: to the blending warehouse and to the conveyor to the enrichment building;
- blending warehouse;
- conveyor to the enrichment building;
- enrichment building;
The color of the final output of one stage corresponds to the color of the input of the next stage.
Dotted lines indicate downtime.
Blue is the original ore input. Red is the final output.
The ore supply is constant, but with a periodic decrease in volume.
Summary.
By nodes.
Breaks in the operation of the crushing and transfer unit with a constant supply of ore.
Summary
By nodes
What a language does is connect certain components, resulting in some meaning.
In the concentrator model, the data and actions performed in the nodes are taken from the
corresponding digital assets. You can take various data and implement valid actions. The result (meaning) of the structure being formed is the volume of the resulting product (time series visualized in graphs).
14. Symbolic paradigm of digital transformation.
After the stage of introducing new information technologies, most companies have accumulated big data on their activities and have everything they need to significantly improve their own business. BigData is not a fiction: the amount of data in companies is huge and a lot of money is spent on data storage. The problem is that it is not entirely clear how to use BigData: where to start, what data is interesting, and what is "garbage".
The barrier to entering a new status is associated with the procedure for determining which tasks are of interest to which data and which are not. It is important to understand the way to solve technical issues: how to get the necessary data (middlware), where and in what form to store (access speed) and how to work with them (data structure transformation and operations).
The digital transformation phase coincided with the rapid development of the symbolic paradigm. In fact, the marketing concept of "digital transformation" adequately reflects the innovations associated with it, if under "digital" we mean the use of abstract symbolic structures (what else is a number?), And under "transformation" - algebraized operations (actions) that do not deduce from the original spaces of meanings when everyone is talking about the same thing.
To ensure that digital transformation is not reduced to a continuous cycle of trial and error,
it is necessary to learn how to formulate business intentions and fully assess their feasibility and effectiveness.
Business intention is an abstraction: you need to clearly describe what does not exist. Assessing intent is a complex computational procedure on real assets and in anticipated circumstances that must be understood and enforced.
The symbolic paradigm allows you to describe what result should be achieved and find options, based on the application of which rules this result can be achieved. The best feature of the symbolic paradigm is that one can consistently create more and more general abstractions at different logical levels and believe that they can be combined into a single whole.
The difference between analytics and the use of the symbolic paradigm is significant. Analytics shows that everything is complex and everything is connected with everything, and the symbolic paradigm allows conclusions to be drawn that allow one to act reasonably and decisively. In a symbolic paradigm, digital business transformation associated with the transformation of what you have into what you need can be implemented on an ongoing basis.
In practical terms, the toolbox of the symbolic paradigm can be thought of as what the entire modern Wolfram Language is.
15. Production and commercial restrictions.
In a company, production and commercial constraints (OECs) exist in large numbers, in different forms and at different logical levels. They can be qualitative, quantitative, logical, etc. For generality, event conditions can be added to the PQS that form the context of doing business (markets, courses, growth rates, etc.).
The symbolic paradigm makes it possible to rewrite POC in the form of symbolic expressions (symbolic operations on symbolic objects), regardless of the nature of the conditions. Symbolic expressions can be reduced to a system of bases that connect symbolic objects after identical transformations.
Further, the basis system can be mapped onto an n-dimensional cube of the required dimension. Such a cube includes exhaustive combinatorial combinations of the evaluated objects. It is necessary to distinguish between an OLAP cube, which is a Cartesian product of actual values, and an n-dimensional cube of the vertices of symbolic objects.
Based on the system of bases, not a single n-dimensional cube will be obtained. Essentially, all the resulting cubes reflect different, but equal constructions of the internal structure of the business, reduced to dedicated PQS.
As a result, it is possible to obtain paths connecting any POC objects that correspond to the real connections of objects in the process of doing business (and not just a demonstration that everything is complicated). If the path includes k-dimensional faces, then this indicates the possibility and necessity of optimization by the totality of objects included in the face. If the path goes only along the edges, then it is enough to make only local rational decisions.
Direct linking of objects allows you to correctly form digital assets and digital twins, as well as SSDL language constructs. In addition, real criteria for assessing the usefulness of the data collected and stored in the company appear.