Behind the scenes of the ABB Industrial AI Accelerator

ABB's Industrial AI Accelerator program aims to foster development of artificial intelligence and drive the next level of the industrial revolution.

“The future of innovation is collaboration”, said Philipp Vorst, project lead of the ABB Industrial AI Accelerator and Research Team Manager at the ABB Corporate Research Center in Germany. Through engaging with external partners, such as universities, research institutes or startups, companies like ABB can identify and capitalize on breakthrough technologies or new business models that go beyond the usual and help the parent company find entirely new offerings for their customers.

Today knowledge is widely distributed and no company, no matter how capable or how big, can innovate effectively on its own. Using this “outside-in” approach, where external ideas are brought into the firm’s own innovation process, has been for long a fundamental part of ABB’s approach to innovation, complementing ABB’s own strong R&D bench.

Artificial intelligence (AI), the next technology frontier or as some call it the “Intelligence Revolution”, is one of those areas where collaboration will make all the difference. While AI is still a growing field, in terms of venture capital (VC) investments it is one of the most well-funded categories. Large companies in every industry are trying to integrate AI capabilities into their offerings. According to Stanford University’s AI Index, the number of active US startups developing AI systems has increased 14x since 2000.

Against this backdrop, seven European startups have been selected to join the ABB Industrial Artificial Intelligence Accelerator to explore industrial applications of their concepts and speed up their development, as ABB continues to pursue innovative digital solutions. ABB launched the program earlier this year to foster the development of AI for industry; the selected startups which joined the 4-month program have been paired with an ABB business line or function to benefit from the company’s deep domain know-how and accelerate their growth. The initiative was conducted jointly with AtomLeap, a Berlin-based startup accelerator and a strategic intelligence provider with a focus on high-tech industries.

To mark the end of the 4-month collaboration period, seven startups presented their solutions, which they had developed together with ABB business lines to around 100 participants. Similar to the TV startup show “Shark Tank”, each startup had to face the critical questions of a jury after pitching its solution. The six-member jury, consisting of business and technology experts as well as investors, selected Greenlytics as the winner.

The Swedish start-up develops tools for AI-based energy forecasting for wind-solar, and consumption, as well as decision support for power trading and plant optimization. This solution has as part of the collaboration project between ABB and Greenlytics already been deployed in ABB’s “Mission to Zero” project at Busch-Jaeger in Lüdenscheid, which was recently opened.

“It has been a great experience to work with ABB, a global technology leader for digital solutions for industry, on the future of energy management,” says Sebastian Haglund El Gaidi, founder of Greenlytics. “We can learn from ABB and use the company’s global presence to deliver our own services. This helps us tremendously. Now we look forward to our future collaboration.”

“We have seen strong synergies between the Greenlytics product offering and ABB OPTIMAX from the start,” explains Markus John from ABB’s Industrial Automation business in Mannheim, who has worked closely with his team with Greenlytics over the past few months. “By combining our systems, we could provide a platform for data collection, forecasting, trading and controlling of distributed power plants. Greenlytics’ AI-based predictions are a good complement to our energy management solution.”

Furthermore, SynerLeap, ABB’s innovation growth hub in the Nordics, will support the startups Greenlytics, Vathos and OneWatt for six months to drive further collaboration with ABB and help the startups accelerate and expand on a global market.

“Working with ABB as one of the world’s leading robotics companies has helped us significantly to advance our product development roadmap and test our solution in a new application,” comments Philipp Küpper, CEO of Vathos. The Düsseldorf, Germany-based startup develops computer vision and machine learning for applications in robotics and factory automation.

Netherlands-based OneWatt has been working with ABB on creating an analysis platform with sound data to predict motor health. “Industrial motors will have problems. If not diagnosed in time, they can result in massive losses, due to unnecessary repairs, or even worse, unplanned downtime”, explains Paolo Samontanez, CTO of OneWatt. “In the last four months we have been addressing this challenge.”

Developing industrial AI through collaboration

While the collaboration projects varied across different industrial applications, they all had in common a clearly defined use case for which the startup and ABB together developed a specific solution.

“Partnerships that foster innovation have long been an integral part of ABB’s approach to innovation. The ABB Industrial AI Accelerator continues this strategy by partnering with startups to drive technologies and solutions that help ABB stay competitive. The close link with the business lines, the R&D teams, and other start-up activities is central,” says Jan-Henning Fabian, head of ABB’s research center in Ladenburg.

Dutch Analytics, for example, has worked together with ABB’s Marine and Ports business line on cooling systems in Medium Voltage Drives onboard vessels with diesel-electric propulsion. Ensuring the right level of cooling fluid during operation of the ship is critical. By using data from the ABB Ability™ Marine Remote Diagnostic System installed on ships, the team developed a model that predicts liquid’s evaporation rate and advise the time of next maintenance.

The Norwegian startup Intelecy was using production data from ABB’s control system ABB Ability™ System 800xA to provide automatic data processing, structuring, labeling, and cleaning of raw data from a specific industrial process. The processed data was then used to select, train and deploy industrial machine learning algorithms and provide new insight.

A very different case was Cobrainer. The Munich-based startup focused its use case on improving access to talent. And for developing AI solutions, the right expertise and skills are key. Together with ABB, Cobrainer built and implemented an intelligent Talent Matching App; the app is based on Cobrainer’s Expertise Graph technology that makes it easy and intuitive to create a personal expertise profile to get highly relevant job recommendations.

Together with ABB’s Oil and Gas team, the use case of Cambridge-based startup 8power focused on answering the problems of diagnostics and monitoring of devices by enhancing the energy supply of sensors: They harvest energy for their sensor from mechanical vibrations. This way, AI algorithms can be integrated into self-powered wireless sensor solutions.

____________________________________________________

The seven startups participating in the ABB Industrial AI Accelerator program were:

  • Cobrainer (Munich, Germany) applies machine learning to enable intelligent employee mobility for large and medium-sized organizations, offering automatic expertise profiling and intelligent job matching.
  • Dutch Analytics (The Hague, Netherlands) offers a predictive maintenance platform which enables the development and deployment of asset monitoring applications. Customers in Rail, Marine and Manufacturing industries are already benefiting by predicting breakdowns and increasing asset uptime.
  • Greenlytics (Stockholm, Sweden) provides solutions for energy forecasting and decision making for power traders as well as plant operators.
  • Intelecy (Oslo, Norway) provides tools to analyze production data from the manufacturing and processing industry using machine learning to prevent breakdowns, predict failures, and improve production processes.
  • OneWatt (Arnhem, The Netherlands) has designed a non-invasive, non-contact predictive motor health maintenance system using sound. The data is gathered through their embedded acoustic recognition sensors, also called EARS.
  • Vathos Robotics (Dusseldorf, Germany) deploys computer vision and machine learning for applications in robotics and factory automation.
  • 8power (Cambridge, UK) produces self-powered wireless sensor solutions for industrial plant applications and machine condition monitoring. The startup uses its patented Vibration Energy Harvester to power their sensors.
Categories and Tags
About the author

Victoria Lietha

I am part of ABB Technology Ventures, ABB’s venture capital unit. I have been with ABB for more than 5 years and I am interested in innovation, technology and sustainable development.
Related stories
Comment on this article