Levelling the Playing Field

3 mins read

Artificial intelligence (AI) is evolving daily, expanding its capabilities and improving existing systems.

As such, engineers are increasingly incorporating AI into their day-to-day tasks. But while many engineers have a range of expertise and familiarity with AI tools, transitioning away from more traditional engineering approaches can be challenging.

The industry needs a solution to help broaden the AI community of users and ensure speed and efficiency across projects.

This is the key driving force behind the rise in popularity amongst engineers of ‘low-code’ solutions which require limited coding expertise to build applications and processes.

Low-code solutions provide a bridge between the work of data scientists on one hand, and engineers solving specific real-world problems on the other. As more engineers use AI to cut production times and costs, low-code will become a natural extension of any AI-driven work, making inter-disciplinary teams more productive and effective.

Yet, in order to reap the benefits, it is important that engineers understand the primary applications, best practices and future trends associated with low-code for AI.

Applications & Benefits 

From industrial automation to radar and wireless, low-code has applications anywhere a data-driven approach is required. Its applications stretch throughout the entirety of the AI workflow, covering data pre-processing, modelling, simulation, testing and deployment.

Data labelling and preparation, for instance, is an area that engineers spend a significant amount of time on, and one that inherently lends itself to an efficient point and click approach.

As engineers work to ensure that clean training data is used as input to models, point and click tools can work to pre-process data such as lidar, radar, signals, and images.

For example, Drass, a marine manufacturing company, used low-code image processing applications to provide clean data sets to train their autonomous surveillance systems in challenging maritime environments. Using point and click tools within MATLAB, Drass could standardise and manage wave motion and persistent background changes, allowing the company to complete its project within 12 months.

This approach has two primary benefits. First, the ease of use seen in solutions like point and click apps allows for interdisciplinary teams with varying levels of coding ability to work together effectively. Engineers with varying levels of coding skill can collaborate to tackle the primary issue their team is facing.

With low-code solutions engineers have the capability to run variations of algorithms in parallel, visualizing the impact and trade-offs of parameters with intuitive interfaces. When multiple iterations are necessary, as is often the case with AI, saving time on writing code across multiple tests can be vital. For example, Path Partner, an embedded systems company that develops autonomous vehicle technology, used machine learning approaches for radar-based automotive applications to cut development times. By implementing point and click applications to compare the accuracy of different classification algorithms in parallel, the firm was able to cut development time from five months to one.

At the Cutting Edge

Low-code solutions also help incorporate the most advanced technology and modelling from data scientists into graphical interfaces that can be used by engineers. 

Often research models are not built with model size constraints in mind, with data scientists developing the latest models on high powered desktops. These models are usually computationally intense and pose issues for deployment on low powered edge devices. Processes such as model compression, however, enable engineers to take the totality of that model and use point and click tools to compress those layers to fit onto a low cost, low powered, embedded system using a graphical environment.

By visualising what happens when a model is compressed through apps, engineers can ensure that the model does not lose fidelity so that it may function accurately on low computational power devices.

As data scientists construct the latest models in Python, incorporating them into programs like MATLAB allows for the transfer of the latest innovation to engineers as fast as possible, without the need to learn additional programming languages. 

Considerations Before Applying Low-code Solutions

The speed and cross-collaborative efficiency that low-code solutions enable make it easier to manage multi-component engineering projects, yet by itself it does not make incorporating AI into everyday workflows an easy task.

Engineers must strive to remember that approaching projects with clarity of purpose is still paramount.

While the strengths of point and click low-code tools can be leveraged in data labelling, without knowledge of what to label, those tools cannot be effective. Similarly, low-code solutions need to be able to incorporate the technology that engineers need. If an engineer pivots from implementing classical machine learning algorithms to deep learning algorithms, the low-code solutions that they are using need to be able to reflect that change.  

Software, or code in general, is still vital to the everyday running of business, and the place of low-code solutions is not to remove code but to limit the investment of time and resources that engineers need to make. With the use of low-code tools, engineers are freed up to do more sophisticated work such as driving greater accuracy through increased iterative testing. 

What is the future?

Used across applications and industries from IA&M to electrification, low-code solutions can bring an added layer of sophistication to the work engineers can produce across data types. As interest in the benefits of AI grow in parallel with its wider adoption, incorporating the newest deep learning and machine learning models into applications for engineers will become essential.

The role of AI in the everyday work of engineers is only expected to grow. As more industries and companies incorporate AI into their work, incorporating low-code apps will become a natural extension of any AI-driven work, making inter-disciplinary teamwork more productive and effective.

In doing so, every engineer can be given the tools to best help him or her succeed with the level of coding they feel most comfortable.

Author Details: Johanna Pingel, AI Product Marketing Manager, MathWorks