Feature
Engineering teams are looking to artificial intelligence (AI) to solve a range of regular issues they encounter in their exacting roles. Engineers are using AI to help manage their production processes, monitor manufacturing problems and make better use of data. Here, we look at some examples of how AI is being used to improve the work of employees in the engineering field.
1. Sweco
Sweco, an architecture and engineering firm based in Belgium, has more than 150,000 projects each year focused on creating sustainable cities and employs more than 22,000 consultants for the analysis, planning and design of those projects.
Engineers and architects at Sweco spend much of their time creating documents, overseeing contracts, communicating with customers and team members, preparing legal paperwork, finding regulatory information and managing budgets, all of which takes away time they could be using for more creative work, according to a case study.
Sweco wanted to automate some of those tasks and develop a proprietary version of ChatGPT, so the company turned to Microsoft. Sweco used Azure AI Studio to automate the creation of documents, such as contracts, and enhanced search capabilities, so consultants could find information faster. The company also designed and deployed the internal chatbot SwecoGPT.
“We wanted an AI service to centralize knowledge, automate more tasks and analyze information — one as good as ChatGPT, but we own it,” says David Hunter, head of AI and automation, Sweco.
Results
- Developed an internal chatbot prototype and deployed it to a few hundred employees in one afternoon
- Built a full version of internal chatbot within two days and fully deployed it two weeks after the initial proof of concept
- Saved employees two or more hours per day
2. Rivian
Rivian, a manufacturer of electric vehicles, equips its electric adventure vehicles and commercial vans with internet of things (IoT) sensors and cameras. With more than 11,000 vehicles on the road, those components were generating multiple terabytes of data per day, which proved hard to analyze and hard to share.
Rivian turned to AWS to make its data more accessible and usable, according to a case study. The company selected the Lakehouse Platform, a collaboration between AWS and its partner Databricks, to unify all of its data into a common view.
The tool allowed Rivian to apply analytics and machine learning (ML) to the information, leading to predictive maintenance for its vehicles and smarter development of new products using various AWS tools. The company could gather accelerometer data to look at all of a vehicle’s motions and understand vehicle performance, driving patterns and connected car systems. This not only allowed Rivian to improve driver safety features, but helped it develop autonomous driving systems.
“I wanted to open our data to a broader audience of less technical users, so they could also leverage data more easily,” says Jason Shiverick, principal data scientist, Rivian.
Results
- A 30%-50% increase in run-time performance, leading to faster insights and model performance
- Capability to perform remote diagnostics on vehicles
- Increased platform users from five to 250 in one year, leading to new ideas about applying ML to data
3. Ather Energy
Ather Energy, based in Bangaluru, India, developed a smart electric scooter to help people navigate India’s crowded cities and congested roadways. The company’s initial platform, however, could handle 10 to 15 internet of things sensors, which limited the capability of the scooters, according to a case study.
Ather Energy needed to build a system that could navigate cities with irregular layouts and many streets with no names, roads with potholes and chronic traffic congestion. The company wanted a touchscreen control that could withstand the heat, vibration and dust in city traffic.
The team at Ather Energy turned to Google Cloud and BigQuery, Google’s AI-ready data platform, to design a new IoT platform. The capabilities from Google allowed the scooters to incorporate 43 sensors and let Ather Energy analyze the data quickly. In addition to providing route navigation with Google Maps, the system provided predictive maintenance, structured electricity use for cost savings and alerted drivers to unoccupied charging stations.
“We’re churning out a few hundred megabytes of data every day on every vehicle,” says Swapnil Jain, co-founder, Ather Energy. “With that sort of data and a few thousand vehicles on the road, we’re doing things which would have been impossible.”
Results
- Allowed platform updates monthly instead of twice per year
- Provided insights to keep costs down for both consumers and the company
- Helped add self-correcting capabilities to vehicles
4. Euramax
Euramax, a producer of coated aluminum and steel coils for recreational vehicles, architecture and transportation, can face changes to its production schedule as new products are ordered or a required material is delivered late. This can make planning difficult and decrease productivity. If one order poses a problem, it’s difficult to see what impact that may have on other orders.
Euramax turned to SAS Viya, an AI and analytics platform, to help reduce the guesswork in production planning, according to a case study.
The system collected data from all of the company’s open orders and retrieved new data from various production systems every 15 minutes. SAS Viya analyzed the data to see if there was a threat in any order that could slow down production and looked at what impact the threat could have on the production of other orders. If it detected a problem, it immediately informed employees, who could respond quickly with an informed decision.
“We can now see in almost real-time how much time we lose in a day,” says Peter Wijers, information communications technology manager, Euramax. “If that number increases during the day, we can immediately adjust the production planning of our other orders. Of course, the system will let us know within 15 minutes whether it was the right decision.”
Results
- All current data collected into one overview for entire company
- Weekly schedule predicted and updated at 15-minute intervals
- System could learn from employees overruling its decisions
5. Siemens Digital Industries Software
Siemens Digital Industries Software provides software and hardware to manufacturers and designers in a wide variety of industries, including aerospace, automotive, consumer products and semiconductors. The company wanted to improve how its customers communicate, so, for instance, shop floor workers can work with operations and engineering teams in real-time.
The team at Siemens Digital Industries Software turned to Microsoft to help create an app to enhance customer communications, according to a case study.
Siemens Digital Industries Software used Azure AI to develop an app that uses natural language processing (NLP) for real-time reporting of issues. The app could use informal speech in various languages, automatically create a summary of the problem, and route it to the right design, manufacturing or engineering experts in the language they require. The app allowed customers to respond more quickly and efficiently to reports of defects.
Developers at Siemens really liked how the cloud AI environment operated, says Manal Dave, advanced software engineer at Siemens Digital Industries Software.
“It definitely accelerated our adoption of advanced machine learning technologies, and they have a lot of confidence now for ongoing AI innovation with this solution and others to come,” Dave says.
Results
- Allowed customers to use mobile devices and natural speech to document issues
- Increased collaboration among teams and simplified workflows
- Allowed service technicians to access simulations and receive virtual assistance with detailed repair instructions from engineers
