Atlas GmbH: How AI is being used in practice in excavators and wheel loaders

Atlas GmbH from Ganderkesee is driving the adoption of artificial intelligence in the construction machinery industry. The German manufacturer of compact excavators and wheel loaders uses AI technology for product optimization and process control. The strategy focuses on practical applications rather than generic digitalization.

Predictive Maintenance Reduces Downtime

Atlas is developing AI-powered maintenance systems for its compact excavator fleet. The software analyzes operating data from telematics systems in real-time. The goal is to predict wear parts before they fail. The system evaluates hydraulic pressure, engine temperature, and operating hours. For construction companies, this means: fewer unplanned downtime, shorter service intervals.

The technology draws on historical data from several thousand machines. The AI model learns typical failure patterns in hydraulic pumps, seals, and filters. The operator receives a warning three to five operating days before the predicted failure. This gives them time to order spare parts and schedule maintenance. Atlas compares the solution to conventional maintenance schedules: a reduction of up to 15 percent in downtime is being targeted.

Efficiency Increase Through Intelligent Machine Control

A second focus is on AI-based process optimization for hydraulic excavators. The software adjusts hydraulic flow and engine speed to the current work task. When digging in dense soil, the system automatically increases hydraulic pressure. During light transport operations, it reduces power. The machine operator needs to intervene manually less often.

Atlas is currently testing this function in a pilot fleet of 50 machines. The first results show fuel savings of 8 to 12 percent compared to standard operation. According to the manufacturer, the technology pays for itself after approximately 2,000 operating hours. At full capacity, this corresponds to about two years. For fleet managers with tight margins, this could be interesting.

Autonomy: Atlas Focuses on Assistance Systems Rather Than Full Automation

Unlike Komatsu or Caterpillar, Atlas does not pursue a fully autonomous machine strategy. The focus is on driver assistance systems. An example is automatic leveling for mini excavators. The system maintains constant digging depth without the operator having to constantly correct it. The function uses sensor data and AI algorithms for soil analysis.

Another project involves load distribution in wheel loaders. The AI calculates the optimal weight per bucket and warns of overloading. This protects tires, frame, and drive components. Atlas argues: The driver remains responsible; the technology only supports them. This differs from the full automation approaches of major manufacturers like Liebherr or Volvo CE.

Data Management as a Challenge

The biggest hurdle for Atlas is the integration of heterogeneous data sources. Machines from different manufacturing years provide different datasets. Older models without telematics interfaces are difficult to retrofit. Retrofitting costs between 1,500 and 3,000 euros depending on machine type. This only pays off for machines with at least 3,000 remaining operating hours.

Atlas is working on standardized interfaces for all current product lines. From 2025 onwards, all new compact excavators and wheel loaders are to be equipped standard with AI-capable control units. The additional costs compared to conventional controls are approximately 5 percent of the machine price. For an 8-ton excavator, this means around 3,000 to 4,000 euros surcharge.

Comparison with Competitors

Wacker Neuson focuses primarily on fleet management software for rental companies. Liebherr concentrates on autonomous large excavators in mining. Komatsu and Caterpillar are investing heavily in fully autonomous systems for open-pit mining and large projects. Atlas positions itself as a pragmatic middle ground for mid-sized construction companies.

The Ganderkesee company focuses on rapid implementation rather than long-term research projects. All AI functions are to be available in series machines within 18 months. This is significantly faster than large corporations, which often take years to bring products to market. Whether this strategy works depends on market acceptance. Many construction companies are skeptical of complex software.

Practical Viability Decides

For the operator, the bottom line is what matters: Does the AI save more than it costs? Atlas calculates annual savings of 500 to 800 euros per machine through fuel and maintenance optimization. For a fleet of 10 machines, that's 5,000 to 8,000 euros per year. The investment in telematics and software pays for itself after two to three years.

Critics complain about dependence on manufacturer software and data access. Atlas therefore offers an exportable data interface. The operator can transfer machine data to their own systems and work with third-party providers. This increases flexibility but requires IT expertise in the company. Whether the AI strategy will prevail will become clear over the next two years.

Atlas GmbH is taking its own path between high-tech promises and construction site reality. Focusing on assistance systems rather than full automation could prove to be an advantage. Those who invest now should calculate carefully: Payback depends heavily on operating hours and usage profile. For fleet managers with high machine utilization, a close look at the solutions from Ganderkesee is worthwhile.