Java 17

 

Java 17 was released in September 2021. This is the latest release of LTS (Long Term Support). This means that this version will receive updated until 2029. An earlier version of LTS (Java 11) was released in September 2018, and between these two there were new releases every six months, but not covered by extended support.

Java was developed slowly to version 8. This appears to be due to a conservative and defensive strategy for introducing changes to the language. At a similar time to Java, the C# language was created, which was characterized by introducing additions on a regular basis to keep the language attractive and effective for programmers. Currently, the Java language is quite often extended with new functionalities but with the usual caution. We’ll cover the latest improvements between versions 11-17 in this article.

 

 

Switch modification

 

Switch in previous versions of Java looked like this:

Source: Learn Java 17 with IntelliJ IDEA, https://www.youtube.com/watch?v=FP0V98S4l9w&ab_channel=Java

This function is used to consider different variants of a value. Depending on the value we want to convey the appropriate information to the user. Of course, we don’t have to use Switch for this task but the Switch syntax makes the code much more readable.

As you can see in the image above, there is a break command after each sequence. If we don’t locate this command in a specific case the executed code will go to the next sequence. Such a mechanism allowed to handle several cases with the same code. This style is also derived from the C language, where it referred to the assembly programming method (a low-level language specific to a given family of processors). Over the years, the style of programming has changed significantly and this use of the break command, or the lack of it, is less and less used. If someone forgets to add the break commands, the code may not work as intended. Therefore, an arrow notation was invented that implies break in arrows. This makes the Switch itself more readable to the user. This is important because there is less chance of making a mistake by accident. Below you can see how this record looks like.

It is also possible to implement a return value, depending on the specific variant. This is advantageous because we can get a result that will be analysed further by the program. These features make the expression Switch much more convenient to use. So developers can write cleaner code that is easier to understand.

 

NullPointerExceptions

 

When programming in a language like Java, there is often a problem with no reference to an object. In earlier versions of the language, the system provided relatively accurate error information but this was not always sufficient to quickly identify the location where the error was causing the error. So far we have been able to find out that it exist at a specific line in the given file. This was problematic because there could be complicated code in one line, which made locating the mistake cumbersome. Before Java 14 the error information was as follows:

Exception in thread „main” java.lang.NullPointerEsception

pl.inero.ExamplesTest.NotMeaningFullNullPointerTest(ExamplesTest.java:155)

Java 14 introduces a facility that makes messages convey more precise information to programmers. This way, programmers can locate the error faster and go straight to repairing the code. An example of information about the lack of reference to the object in, after the implementation of this improvement, looks like this:

Cannot invoke „String.toUpperCase()” because „testOne.company” is null

 

 

Improvement to String constant in Java 17

 

Java 15 introduces so-called Text Blocks. Thanks to them, entering multi-line fixed string is much more convenient. Before Java 15 you had to enter multi-line constant strings on one line and save newlines as /n. The only way to write a multi-line constant in the code on many lines (for readability) was to divide it into smaller fragments connected by the + operator.

In java 15 start a block of text with three quotation marks “ “ “ and a new line. All subsequent lines up to the next sequence of three quotation marks “ “ “ will be the contents of the constant. It is not necessary to add newlines with /n. In addition, the indents will be left-aligned as much as possible.

Multi-line String constant before Java 15:

String beforeTextBlock = "SELECT some_id, COUNT(*)\n" + 
        "FROM (\n" + 
        "         select da.some_date dd, d.id some_id\n" + 
        "         from some_aud da\n" + 
        "     ) as subq\n" + 
        "GROUP BY some_id\n" + 
        "\t\tDouble \"Double \"quotes\"\" test\n";

Equivalent multi-line constant String after the introduction of Java 15:

String withTextBlock = """ 
        SELECT some_id, COUNT(*) 
        FROM ( 
                 select da.some_date dd, d.id some_id 
                 from some_aud da 
             ) as subq 
        GROUP BY some_id 
        \t\tDouble „Double „quotes”” test 
        """;

 

Pattern Matching

 

Java 16 introduced pattern matching extension for instanceof operator. Before introducing this functionality, after successfully verifying the class of an object, in order to use its properties, you had to manually cast the object to the verified type. It looked like this:

if(someObject instanceof String) { 
    String someString = (String) someObject; 
    //Operacje na zmiennej someString 
}

However, thanks to pattern matching for instanceof we are able to express the program logic more clearly. Below you can see what the code looks like:

if(someObject instanceof String someString) { 
    //Operacje na zmiennej someString 
}

Thanks to this update, we can check if an object is an instance of a given class and define a new variable that we assign to the Spring type with one notation:

someObject instanceof String someString

This provision:

1. Checks if someObject is an instance of the String class,
2. cast someObject to the String type,
3. and assigns a String object to the variable someString.

 

 

Records

 

Records have been introduced in Java 16. They allow easier creation of classes for data models. Records automatically implement equals(), hashCode(), toString() and field accessors.

Thanks to this, instead of describing the class with data like this:

public final class NoRecordClass { 
    private final String title; 
    private final String author; 

    public NoRecordClass(String title, String author, String isbn) { 
        this.title = title; 
        this.author = author; 
    } 

    public String getTitle() { 
        return title; 
    } 

    public String getAuthor() { 
        return author; 
    } 

    @Override 
    public boolean equals(Object o) { 
        if (this == o) return true; 
        if (o == null || getClass() != o.getClass()) return false; 
        NoRecordClass book = (NoRecordClass) o; 
        return Objects.equals(title, book.title) 
                && Objects.equals(author, book.author); 
    } 

    @Override 
    public int hashCode() { 
        return Objects.hash(title, author); 
    } 
}

We can use such a short notation:

public record RecordExample(String title, String author) { 
}

Note that for a Java 16 record the accessors differ from those shown in the previous sample – to get the value corresponding to the title field in NoRecordClass use the getTitle() method. However, for the RecordExample class use the title() method, that was generated at compile time. As you can see, the difference is that the accessors generated for the record do not have the prefix get. For this reason, the records do not follow the Java Beans encoding convention. According to the JEP 395 document, the records are primarily to play the role of the so-called named tuple rather than the easier creation of immutable Java Beans. The difference is rather subtle but it can make it incompatible with some older libraries.

 

 

Sealed modifier in Java 17

 

The last functionality we describe is especially useful for library creators. It is used to control what classes/interfaces can inherit or implement a given class/interface. Thanks to this, no unauthorized person can connect his class to our library. Below you will find what such a record looks like:

public sealed interface SealedInterface 
        permits ImplementationNo1, 
                ImplementationNo2, 
                ImplementationNo3 { 
    int getNumber(); 
    String getText(); 
    List<String> getStrings(); 
}

 

 

Summary

 

Java 17 doesn’t actually provide any spectacular modifications. In fact, it stabilizes the changes that occurred during version 12-16. However, we can point out a few fixes that will definitely improve the work of programmers. This is primarily about: records, switch, pattern matching, sealed modifier or text blocks. Java 17 is definitely a big step forward, mainly because it is an LTS version, Thanks to this, developers will be able to count on ongoing support for the functionalities mentioned in this article.

Inero Software provides knowledge and expertise on how to successfully use cutting edge technologies and data to shape corporate digital products of the future.

In the blog post section you will find other articles about IT systems and more!

Consulted by: Tomasz Bieliński.

 

How to build your own custom voice command recognition system?

 

The concept of voice assistant was created to improve human life. It plays the role of an tool which will find appropriate information for us, facilitate online shopping or manage the operation of various devices without any pressing or regulating. However, creating such a system from scratch can be a difficult challenge. Especially if we want to build it in language other than english.

There are general models of neural networks that are dedicated to spoken language recognition, i.e. transcription from audio to text. Everything depends on how much data we have. Nowadays, there are many recordings of the English language that are decently prepared. However, data alone is not enough. Transcripts must be attached to them. Unfortunately, in Polish we do not have a many voice samples with transcripts. So today we are not able to teach models as well as it is done in English.

So far, we cannot train assistants using open source models with satisfactory effectiveness, but we can check what we are able to build from generally available data. The goal of our work is to create a speech recognition system that actually works.

 

 

Preparation and standardization of data for voice command recognition system

 

First, we need to ensure the minimum and maximum length of the recording. They must be within a certain time frame. Long recording may be a problem when training the model, while short recording may not carry any relevant information. We also need to ensure that the signal is sampled at the same rate. This is an important point when we are teaching sound based models.

A pure tone sound wave propagating in space is sinusoidal. Changing the frequency of such a sound means a change in the period of the sinusoid, so the distances between its crests change. However, the sound in the computer also has to be presented in some way. The sine wave is continuous and the computer stores individual values. The sampling rate tells us how many times per second the value of the sound wave was recorded. It is important that each sound that enters our model has the same sampling rate.

Due to incorrect sampling the reconstructed sound can be distorted or completely inaudible. Therefore, it is necessary to remember about the appropriate selection of the sampling frequency to minimize the size of the saved data with the simultaneous lack of information loss. Sampling too rarely can cause aliasing, i.e. the overlap between higher frequencies and lower frequencies. For example, a sound that is a mechanical sinusoidal wave with a frequency of 3 Hz may be misinterpreted as a lower frequency signal. This situation is presented in the figure below, in which a sinusoidal signal with a frequency of 5 Hz was interpreted as a signal with a frequency of 1 Hz due to a too low sampling frequency.

Source: Reprezentacje danych dźwiękowych w kontekście metod uczenia maszynowego, Tymoteusz Cejrowski

This is prevented by sampling the signal at a frequency of at least twice the highest frequency in the signal. This is called the Nyquist frequency. For example, recording on CDs are saved at a sampling rate of 44100 Hz. On the other hand, the maximum correctly reconstructed frequency will be 22050 Hz, which corresponds to the upper range of sounds audible by humans.

 

Description and teaching of the deepspeech model

 

Neural networks are somewhat like the nervous system. The basic units in such networks are neurons, which are arranged in layers. It is a simplified model of information processing by the human mind.

Typically, a neural network consists of three parts. The first is the input layer which has units representing the input variables. The next part is the hidden layers that contain the unobserved entities and are the hidden state of the neural network. It is on these layers that the efficiency of the model largely depends. The last part, of course, is the output layer, which has units representing the targets. All interconnected units are specific connections of different weights. The input data is on the first later, it goes through the next layer, and finally we get the result from the output layer.

The neural network can learn by comparing records. It generates predictions for specific data (audio recordings) and introduces weights adjustments if the generate an incorrect prediction (wrong transcription). The whole process is repeated many times until the accuracy is satisfactory. All weights are random at first and outgoing responses don’t make such sense, while the network improves its predictions over time. This is because the network adapts to the data as it learns.

One of the simplest models of Neural Networks is the Multilayer Perceptron. It consists of many neural layers. The neurons of the previous layer form a specific vector that is fed to the entry of neurons into the next layer. A single neuron in the next layer has the number of inputs equal to the number of neurons in the previous layer +1. However, within one layer, the neurons do not have any connections with each other. This type of model is also known as the “Feed-Forward” model and is one of the fundamental architectures of neural networks.

Figure representing the “Feed-Forward” model.

Recursive networks differ from Feed-Forward networks in many respects. One of the main differences is how the input data is propagated to the output of the network. In the simplest model, the network output is nothing but a combination of weights and the model input (single record). For recursive networks, the model output also depends on the previous output. The result of the model operation for the X record is taken into account when calculating the output for the next Y record.

A convolutional neural network is the type of network most commonly used for visual image analysis due to its input processing nature. Like a human, this network does not analyse the image pixel by pixel, but captures the patterns present in the input thanks to the use of the so-called Kernels or filters. Each layer in the neural network learns image characteristics such as contours and light saturation. Submitting this information improves the efficiency of the model.

However, it is worth looking at open source systems that enable the creation of voice command recognition. One of them is Kaldi, which was written in C++. It was created in 2009 and its main features are that the system is extensible. The community itself provides many other modules that you can use for your own tasks. Kaldi also supports deep neural networks. Although it is written mainly in C++, it has Bash and Python scripts.

For the construction of the voice command recognition system, we focused on the deepspeech 2 model. It uses deep learning and it consists of 3 convolutional layers, 8 recursive layers and one Fully-Connected (described as Feed-Forward). Deepspeech 2 can be successfully trained in any language, but you have to remember that models are most often issued only with code, so you have to learn them yourself. For popular languages such as English and Mandarin, there are pre-trained ready-to-use models. For our assignment, we chose deepspeech 2 with a PyTorch implementation because it is lightweight, relatively few parameters (weights to learn), and is actively maintained. In this model, in addition to preparing the appropriate sampling frequencies about which we wrote above, we also had to prepare the data for the deepspeech model itself. This was related to the provision of an appropriate directory structure.

 

 

Reasoning with the use of ASR models and an additional language model

 

Using deepspeech we are able to apply an additional language model (n-gram model). It is based on statistics and helps predict the next element in the sequence. It must be remembered that the application of such a model requires the collection of a large amount of statistical data. N-grams help machines understand a word in a specific context. This allows them to understand its purpose better. So if you want to build an effective voice command recognition system, it is worth considering.

If we want to create such a model, we start by counting the occurrences of a sequence of a fixed length in the existing linguistic resources. Therefore, whole texts are analysed and single occurrences (1 gram) twos (2 grams) and triples (3 grams) are counted. We can also obtain a 4-gram model, but here huge sets of data are needed, which makes it challenging for the Polish language to implement.

Then you convert the number of occurrences into probability by normalizing. In this way, we obtain the prediction of the next element based on the sequences so far. The more text we analyse, the higher the quality of the model will be. This data is used in natural language processing (NLP) applications. However, there are methods that allow n-gram models to be improved by smoothing the collected statistics. This model is primarily characterized by scalability. We can choose a schema without a lot of data, but then it will be less predictive. The n-gram model for the Polish language can be found here.

The use of the n-gram model at the inference stage allows to correct the output of the neural network (transcription of the deepspeech 2 model) in accordance with the rules contained in the n-gram model.

 

 

Operation on a limited set of commands

 

Unfortunately, the models built cannot be learned well enough due to the small amount of data in a language we work in – Polish. In our case, we can operate in a limited set of commands. This means that a specific number of voice commands will be recognized. Our task was to match the appropriate command from the set with what the model provides us. It is important here to define the highest similarity between the command and what the model returned to us.

For this task, we used a metric called Levenshtein distance, which indicates the similarity between the transcript and a given command. It consists of counting items or letters that do not match. For example, the Levenshtein distance between the terms:

• Gravity
• Gravity

Is zero. These words are identical, so no action is required here.

On the other hand, the Levenshtein distance between the words:

• Long
• Pork

Is 3, because it takes at least 3 actions: replace L with P, replace n with r, and replace g with k.

Where there are fewest substitutions, we have the greatest probability that the commands are the same.

 

 

Summary

 

The purpose of voice recognition systems is to ensure ease of communication between the device and the human being. Speech recognition technology is already a part of our daily lives. For now, it is limited to relatively simple commands. Building such a system based on data in the Polish language is a difficult task due to the small amount of data. Taking into account the current trends in the use of smart devices and technical possibilities, it is worth exploring the potential of these systems.

Inero Software provides knowledge and expertise on how to successfully use cutting edge technologies and data to shape corporate digital products of the future.

In the blog post section you will find other articles about IT systems and more!

Consulted by: Tymoteusz Cejrowski, Software Developer.

 

Working from remote locations, recently has became a standard in many companies. During past few years we have learnt to utilize most of our IT systems, programs and tools remotely such as Office, Outlook, Microsoft Teams, Google Meets, Google Work Space or Google Docs, documentation flow systems or others. Due to significant physical distance to used IT infrastructure, the access to these tools must be properly secured and configured to allow efficient use but third parties, suppliers or other employees have limited access to what you do, in accordance with company’s security policy. Such access is most often implemented in programs through a system of individual accounts in which everyone uses individual logins and passwords.

In large corporate teams with employees working in different locations, this issue is even more important. Although people employed in such teams work in different countries or time zones, many resources must be synchronized with each other and unified access to them improves work, facilitates communication and team management.

Examples of such systems include e-mails, email responders, instant messaging, invoicing, document management systems and many more. In this context, problems appear when the same person needs and access to two or more systems, for example, a contract management tool and document flow management systems. While having many similar situation, hierarchical privileges managements, various systems, remembering several logins and passwords may be confusing for some employees. The desired scenario in such case would be to have on logging gateway that would allow for access to different resources.

Here comes digital identity management!
When an employee logs into the companies resources, he wants to have one digital credentials (i.e. login & password), because in fact he represents the same person in each of these systems. He is also a subject to a specific group of authorizations and rights resulting from the assigned position, duties and other. In such cases Identity and Access Management (IAM) are a perfect choice!

Digital identity is a collection of individual credentials (mostly logins and passwords) that as users, we use in various structures. When we use many systems, we have a lot of credentials. In the long run, this can be troublesome, especially in situations when we forget one of the passwords, because we will need to recover them later. In this case, it is difficult for both users and administrators to manage it.

IAM is a tool that provides effective management of access to information resources. With this solution you can handle the requests of newly hired employees. The identity and access management system also helps in administering users who have parted ways or changed their status with company. With IAM changing or expiring their permissions is much easier.

One of the most frequent concepts in the development of IAM is the concept of SSO (Single-Sign-On). It’s a set of systems in which we have one common login point and one common module for managing users and their authorization to all systems. They allow us to manage digital identity in one place. We have one login and password for many systems to which we can log in with the same credentials.

 

Impact of COVID-19 on Integrated Authorization Management 

 

In the last two years, IAM systems have been increasingly used due to the need for remote or hybrid work. IAM is of utmost importance in supporting and securing digital resources. Tools like this provide basic authentication and authorization to secure employee data. This goal can be achieved through access management (Multi-Factor Authentication), enabling enterprises to provide end-user authentication to multiple systems. 

 

 

When is Identity and Access Management important? 

 

Digital identity management actually always happens when we use our unique login. A simple example of using a digital identity is integrating a Google account with the Chrome browser. In the case of corporate solutions, our identity must be better authenticated, so each time we log into the system we have to enter our login and password. Digital identity management occurs when we want to have access to information that is intended only for us, or it requires our interactions.

“According to a report by Verizon Data Breach Investigation, more than 70% of employees re-use their passwords at work. The report found that 81% of hacking breaches used stolen or weak passwords.”

 

The basic principle of creating IAM systems is not to create new security solutions but above all, to use standards and proven security methods instead. Tools that allow us to use standard approaches to ensure the safety and convenience of use in authorization for various connection scenarios of various systems, including:

• Azure B2C 
• OAuth2 
• Open ID Connect 
• Identity Server 
• Active Directory 
• KeyCloak 
• Amazon Cognito 
• Google IAM 

When mentioning digital identity management, it is necessary to indicate the differences between authorization and authentication. Authorization allows access to confirm whether a given person is allowed to use a resource (e.g. a function or database), and authentication means confirming the identity. 

 

5 benefits of IAM

 

Increasing cyber security 

Thanks to IAM solutions, companies can implement security policies in all connected systems. Administrators using such tools can easily remove unwanted access permissions when needed, by providing one consistent system of accounts and passwords. 

Lower operating costs in terms of infrastructure management and IT security 

With integrated systems enterprises can benefit from lowering costs in IT infrastructure by minimizing the time needed to solve problems related to the user account.

More convenient use of systems by users and administrators 

By implementing IAM tools, administrators are able to create a unique identity for each user. They don’t have to manage dozens of accounts for different applications or other resources. Users have access to systems regardless of their location, time or device currently used. 

Easier adaptation to the regulations of corporate security policies 

When ordering or creating software, corporations implement security policies about what can and cannot be done in specific IT systems. If we are dealing with one account logging system, it is much easier to adapt it to security policies. 

Easier password management and recovery 

With IAM solutions, password problems will be minimized. They help administrators implement better password management practices. We are talking about frequent updates of login credentials or stronger authentication. 

 

 

Identity and Access Management 

 

If enterprises want to keep their employees safe and increase their productivity, they should opt for integrated identity and access management. After logging into the main system, users don’t have to worry about having an appropriate password for other structures. The employee has access to the perfect set of tools to increase his productivity.

“72% of organizations prioritize security over operational efficiency (52%) and tamper prevention (47%) as key factors in the development of the IAM program.”

~According to 2020 IAM Report, Cybersecurity Insiders

 

Digital identity management is a process that can be implemented in stages. It’s not necessary to integrate all systems at once. Company may decide to implement single systems and add more structures over time. 

Inero Software provides knowledge and expertise on how to successfully use cutting edge technologies and data to shape corporate digital products of the future. In recent months we have implemented several cybersecurity solutions based on IAM that allow users to use single sign-on authorization point and securely access corporate systems.

In the blog post section you will find other articles about IT systems and more! 

 

 

Huge amounts of data in business life 

 

With the revolution of Industry 4.0, production became automated and more effective. We have more automatic devices, precise measurements, storage capacity and computing power that makes the information easier to handle. The huge amounts of data is becoming the driving force of industrial production and cover all its aspects. Data can be about products, equipment, research, supply chain, finance, users or even their behaviour.

Nowadays, data is collected by automated devices to manage production effectively and processes. Thanks to the appropriate data flow, the modules included in the production process will be more and more refined. In this way, each part of the production process becomes more detailed, allowing the company to locally adjust production, reflecting and anticipating the needs of customers. Therefore, an appropriate product cycle is created: production -> sales -> feedback. 

The amount and complexity of the data make their interpretation and adoption beyond human means. With a complex production process, it is challenging to tell which machine is responsible for the degradation in quality or performance. The same applies to the quality of the goods if the speed of their production is enormous. Using integrated data processing, we can divide the modules into various tools, such as data transmission systems, data acquisition systems or data feedback systems and therefore virtually model their behaviour and optimize it’s performance.

 

 

Business process optimization 

 

However, data alone is not enough. It is important to extract valuable information and knowledge from the data. That’s what Machine Learning does. It allows you to assess trends, relationships between data, relationships between measured data and other data from various processes or predict events based on what has happened. Manufacturing plats today are becoming more and more complex with interrelated processes. In such situations, a person quickly reaches the limit of his ability to take into account every aspect of the process. Machine Learning offers an effective solution that overcomes the challenges mentioned above.

“As companies attempt to increase productivity within research, chip design, and manufacturing while simultaneously accelerating time to market, AI/ML is becoming an increasingly important tool along the whole value chain.”

~According to McKinsey

If we want to extract information from data, we need to perform the data extraction process, remove unnecessary dependencies, clean gross errors, understand the existing dependencies, develop and adjust the ML model that will map the data and finally adapt it to work in production mode. 

 

 

Machine Learning areas of applications 

 

The main purpose of Artificial Intelligence algorithms is to solve specific problems and improve processes. Machine Learning allows you to optimize areas such as: 

• Process automation 
• Work safety 
• Quality control 
• Decision making

It is worth mentioning that algorithms allow you to solve problems from various business spheres. Therefore, it is necessary to determine in business what Machine Learning approaches can be applied in our company. We should set ourselves up for the fact that implementing ML is a process, not a task. The implementation of machine learning algorithms requires constantly generated data to be able to control business processes properly. Therefore, all aspects of such processes must be digitized as much as possible. Therefore, it requires experts to design processes taking into account the current state of the data, thanks to which it will be possible to collect correct data consciously and optimize the appropriate metrics.

“McKinsey predicts that Machine Learning will help manufacturing companies reduce material delivery times by 30% and achieve fuel savings of 12% by optimizing their processes. The company also estimates that enterprises can increase gross revenues by 13% if they fully integrate AI technologies into their operations.”

Machine learning algorithms can run in real time, so that they will provide specific recommendations or they identify potential for improvement in production. Machine Learning can therefore serve as a support tool for process control operators, helping them make more informed decisions to maximize production in their enterprise. 

When implementing solutions in the area of machine learning, we must build and appropriate strategy in line with the goals of our company. By defining clear goals and understanding the required infrastructure, we can count on a greater probability of success of the implemented solutions. In addition, it must be remembered that the data must be properly formatted and organized in terms of the algorithm. If we want to take full advantage of machine learning in our organization, we need an appropriate team. By employing passionate people, the probability of success will increase dramatically. The Inero Software team offers AI / ML implementation to enterprises in a few steps: 

1. Identification of problems and business goals – this is crucial when trying to understand a given infrastructure.
2. Establishing data sources and their quality – the better quality of the data we have, the more effective machine learning algorithm will be.
3. Data collection and adaption – adjusting the data and organizing it will also positively affect the operation of the algorithm.
4. Creating a model based on points 1-3.
5. Launching and verification in the context of the assumed goals.
6. Establishing bottlenecks where we have data gaps – so we check what data is badly collected and draw conclusions from it.

In this way, we get an effective model that efficiently analyzes the data and provides us with accurate results. Of course, building such a model is a continuous process, so it is worth repeating the above steps from time to time to further improve our algorithm.

 

 

Effective enterprise management with business process optimization

 

It is important to understand that Machine Learning is not the best solution to all problems. ML is a tool that allows you to effectively use data to manage and optimize processes. Many industries currently focus on digitization and analytics. In some businesses, huge amounts of data accumulate every day and actually, machine learning algorithms can collect and analyse an unlimited amount of data. On the other hand, the optimization performed by experts is largely based on their experience and knowledge about the process itself. Machine Learning is about learning from previous these previous experiences. So by analysing such amounts of data, algorithms can learn to understand the complex relationships between different business subcomponents and their impact on final result. How AI learns from experience is in a way similar to experts who learn to steer a given process. However, AI allows you to analyse complete data sets over several years. 

The cooperation of Machine Learning experts and business experts is important in this matter so that ML reflects the business processes. On the one hand, what counts here is a good specification of the ergo business needs and the optimization of ML for what we want to achieve. On the other hand, it is important that business experts understand the limitations and disadvantages of Machine Learning. Bad data will cause the applications to end up damaging the process. 

Inero Software provides knowledge and expertise on how to successfully use cutting edge technologies and data to shape corporate digital products of the future.

In the blog post section, you can find other articles on modern enterprise solutions. 

 

 

In recent years, there has been a significant development of services and products from the MedTech industry. The digitization of services allows for a better use of resources (funds, care, medicines) and people (shortening the travel time, fewer documents). Thanks to this we can focus on the effective way to improve caring for the elderly. Digital transformation gives a start to many tools and resources which improve providing such services. 

In this blog post, we present our experience of working on the application, which we developed in cooperation with the Social Services Center in Warsaw and implemented in July 2021. This is a system that enables management, planning and verification of hot meals delivery by the administrative unit for people in need.

 

Reporting and deliveries analysis – problem description

 

Systematic hot meals delivery for elderly or impaired people can constitute a significant logistical and organizational problems for social case services. Usually hot meals are delivered in lunch time, for instance between 11am to 3pm and the food is delivered to the specific premises where a given person lives. Meals are often not received due to a person’s departure or health problems. The social services center usually employs catering companies to prepare and deliver meals. They must have confirmation that the food has been delivered to the resident on schedule. Without a proper system, controlling meal deliveries can be problematic.

In care services it is important to create constantly updated lists of people to whom hot meals are delivered. More over, the it must be taken into account that certain decisions determine whether a particular person is entitled to a meal on a given day. For some people, meals are delivered daily, for others on weekdays, and for others only on weekends. This makes the daily preparation of an updated deliver  list time-consuming, and the activity itself is repetitive and error-prone.

Overlapping organizational and billing problems in care services that supports a large number of residents everyday, makes the process not completely successful. Digitalization improves organizational and logistics processes. Thanks to this, people who need help actually receive it.

 

Benefits of digitalization in social care services

 

Digital transformation improves the whole process associated with the cooperation of entities. On the one hand, we have a town hall that orders the execution of meals, and on the other hand, catering companies and suppliers who prepare and deliver meals. At the end of this process, we have people who need help.

Our application automatically generates meal delivery lists for each day based on the entered and existing administrative decisions. Thanks to this, we can be sure that every person that needs help will receive their meal. 

Digitalization allows for ongoing verification that hot meals are properly delivered. For this we can use QR codes. Stickers with such codes are placed next to each apartment. Suppliers can scan them and confirm that on schedule, they arrived at the location and delivered the meal.

Using mobile app you can inform town hall that a specific person has not turned up at his/her place of residence. Proper system allows you to automatic generation of reports and settlements for catering companies. Information about delivery times or delivered meals are contain in the server part of the system, to which town hall employees have access. Thanks to this, they can better control spending money from public funds.

IT technologies ensure that people in need of help receive their hot meal on time without any delay. Using mobile apps employees of catering companies and town hall have access to up-to-date data. CUS Warsaw uses such solutions to deliver meals for elderly. 

 

Digitalization of social meal delivery for elderly and impaired people 

 

We created a system which consists of a mobile application and a server part. It allows to management, planning and verification meals delivery to people who need help. This project supports social policy through which people who need care can receive it. In every large city there is a center dedicated to this, which deals with care services. For example, in Warsaw it’s Social Service Center “Social Warsaw”, and in Gdansk its Municipal Family Support Center.

 

 

The main assumptions of a project 

 

The main task in this project is the automation of supply chain management using digital technologies and reporting. System enables creating of updated delivery plans for each day. It automatically generates daily and monthly reports. The system can handle up to 2 thousand orders a day. Technologies that are used in this project enable automatic management of people covered by delivery, catering companies and drivers in the system. It processes personal data, but it’s compliant with ISO 27001 standards.

Thanks to this, all data is safe. Only the required data is processed, so it is not distributed to the outside world. So, this project meets all GDPR requirements. Our system ensure insight into the current implementation of delivery processes and informing about the unavailability of a resident. It provides up-to-date information on meal distribution, i.e. the type of diet, transporting times and residents comments.

 

 

Advantages of implementing the system 

 

Implementation of this project makes the entire delivery process transparent. It can shorten the time of administrative and logistic services using technological tools. If the process is more transparent, the users of local government administration have better insight into how public money is used for social assistance. Thanks to this system, obtaining reports and summaries is much more convenient and less time-consuming. The entire process of spending public funds is transparent to all parties.It integrates systems that already exist in CUS Warsaw. Therefore, importing various types of data from existing IT systems is much easier.

 

 

Improving social care services 

 

The use of IT technology in the care services makes the entire process their provision much more transparent and, above all, optimized in organizational, formal and financial terms. Thanks to this, both town hall employees and catering companies are gaining on time. This is a significant factor when delivering hot meals to people who need help. Our system is a cloud-based system, but it can also be implemented as an on-demand system. So, it can be used in the infrastructure of the office or in dedicated servers.

The implementation of such a system is not a complicated process. You have to take care of:

• Integration with the already existing systems or with other systems for decision about delivering meals, 
• Delivering QR codes to residents, 
• Installation of the application by drivers. 

As a company, we offer effective solutions that facilitate the effective management of logistic and organizational processes. We help in monitoring and reporting activities in the supply chain. 

Inero Software provides knowledge and expertise on how to successfully use cutting edge technologies and data to shape corporate digital products of the future.

In the blog post section, you can find other articles about our projects and more!