Lightly launches edge AI tool to boost European car autonomy

Swiss artificial intelligence startup Lightly has introduced a new edge-based software solution aimed at addressing the data collection and processing challenges faced by European automotive manufacturers working on autonomous vehicle technologies.

LightlyEdge, the company's latest offering, is designed to optimise the way vehicle cameras and sensors gather information by deploying intelligent models that analyse data in real time and filter out redundancy at the moment of capture.

This development emerges as European automakers contend with significant competitive and economic pressures, including rising tariffs, talent shortages, and rivalry from American and Asian firms such as Tesla and BYD.

The need for efficient data collection has grown as the automotive industry pushes to advance autonomous driving capabilities. Collecting vast quantities of sensor and video data is common practice, but distinguishing relevant data from redundant material remains a persistent challenge.

LightlyEdge addresses this issue by determining which data instances—such as unusual or critical events like a child at a crossroad or an accident unfolding in snowy conditions—are most valuable for AI training purposes. By capturing only these uncommon or informative scenarios, the system reduces the storage and bandwidth requirements typically associated with developing machine learning models for driving.

In recent months, European automotive companies have faced the dual pressure of internal workforce reductions and external market competition. The sector has seen manufacturers such as Bosch announce plans to cut 12,000 jobs, while others like Mercedes have offered up to EUR €500,000 for voluntary exits.

Matthias Heller, Co-Founder of Lightly, said, "This launch is a fantastic opportunity to increase development cycles for autonomous driving and driving assistance. With LightlyEdge, our partners can harness smarter, real-time data collection that not only accelerates AI model training but also provides a competitive edge against established industry giants. By focusing on quality data from uncommon scenarios and hazards, we're empowering a new era of innovation that will drive the future of mobility."

LightlyEdge reflects a broader industry shift towards intelligent selection of training data—an approach typified by the "Active Learning" methods popularised by companies like Tesla, which prioritise the collection of rare or instructive examples and minimise data overload.

By deploying their solution directly on edge devices within vehicles, Lightly aims to provide developers with real-time feedback and smart data capture tools, allowing new AI models to be tested and refined on public roads efficiently. The company claims that this approach will significantly reduce the time required for European manufacturers to develop and release new autonomous features.

LightlyEdge builds on the capabilities of the company's existing LightlyOne platform, which operates in data centre environments. With the introduction of LightlyEdge, the technology has been migrated from centralised sources to the vehicles themselves, aligning with industry trends towards edge computing and distributed artificial intelligence.

The user interface is designed for ease of adoption, letting engineering teams monitor performance, deploy updates, and enhance their AI-driven systems with minimal additional overhead.

The automotive sector in Europe has been under increasing pressure to keep pace with rapid developments overseas, amid accusations of lagging behind US and Chinese counterparts in the autonomous and electric vehicle race. Products such as LightlyEdge potentially offer a way for local manufacturers to remain competitive by facilitating faster iterations with lower hardware and operational costs.

Lightly was founded in Zurich by Matthias Heller and Igor Susmelj, who have backgrounds in autonomous driving research and computer vision. The company reports that its technology is already in use among Fortune 500 companies and startups involved in robotics, autonomous driving, and video analytics sectors.

The new edge-based solution is expected to provide European automotive manufacturers with means to adapt more quickly to evolving market expectations and further their investments in the research and development of advanced driver assistance and autonomous vehicle systems.