Siemens Taps into AI to Improve Customer Journeys

Maria Sundvall

The future is data-driven. Cisco predicts that by 2021, every person will have 3.5 connected devices generating 35 gigabytes of traffic per month. Business traffic is also expected to grow to more than 45 exabytes per month in the same time period. [1] People, machines and software are all contributing to an ever-growing volume of data sets that need to move both faster and on an increasingly distributed basis. If businesses are to compete in this data intensive era, they will need to move on from process-driven models and harness the power of data analytics instead. As a result, more and more companies are turning to Artificial Intelligence (AI) to support their business.

Interconnection, or the direct and private exchange of data between business partners is in Equinix’s DNA, and a strategic enabler to digital transformation. Emerging technologies such as AI cannot exist without real-time data flow between a variety of different components. Without interconnection, companies are stuck backhauling all their data between multiple sources and centralized corporate data centers.

I recently participated in Sweden’s biggest AI podcast where I discussed how the new technology is changing society as we know it.

The podcast, named AI-podden (in Swedish only), is hosted by AI expert Ather Gattami, senior researcher at RISE (Research Institute of Sweden), and founder of Bitynamics, a firm that helps businesses become AI-driven. In the 20-minute podcasts, Ather and his guests aim to popularise the topic as well as provide some insights as to how organizations in Sweden are currently making use of the technology.

Predicting failures

One of the subjects I touched upon in the interview was predictive maintenance (PdM), which aims to prevent equipment failures by analysing production data to identify patterns and predict potential issues before they happen. The potential is huge as companies can suddenly benefits from avoiding equipment downtime, reducing costs while increasing efficiency and customer satisfaction.

Equinix is involved in several projects where our customers are developing AI systems. We try to help our customers optimize their IT strategy to handle the type of data loads that AI infrastructures need to consume.

For example, one of Equinix’s customers in the aerospace industry collects data whilst aircrafts are in the air through the aid of sensors. The information is then fed into an AI model which can detect and predict technical faults before they occur – with potentially lifesaving implications for passengers.

Distributed infrastructures

The huge data volumes generated from so many different devices across fleet of aircrafts means that such an AI architecture must be distributed at local edge locations, physically closer to the data sources. This customer in particular had originally built a centralized architecture, but as each aircraft started to collect on average 4TB of data each day, transporting it to a centralized data center quickly became too expensive to build effective AI models. Equinix was able to help this customer deploy an IT architecture at the edge, enabling them to reduce cost, transfer and process data much quicker, and in turn reduce the number of grounded flights due to technical issues through improved predictive maintenance models.

A similar example to illustrate the use of this technology is Siemens – a leader in engineering solutions, who together with Teradata worked to leverage sensor data analytics and predictive maintenance to reduce train failures.

They built a system where more than 300 sensors on each train collect data which is then analyzed together with historical data to predict faulty components. The company analyses sensor data in near real-time, which means that they can react very quickly if there’s an indication of a potential problem, ensuring greater uptime for train operators, fewer delays for passengers and more cost-effective maintenance. Siemens is now able to predict errors three days in advance, reducing overall costs and improving the customer experience. By deploying on Platform Equinix™, Siemens re-architected its IT service management infrastructure for a digital edge by integrating digital and cloud technologies, while optimizing automation, improving reliability and boosting performance.

Predictive maintenance is creating new opportunities for businesses to cut costs and improve productivity – but it can also save lives.

Learn how European transport operators bet on predictive maintenance solutions to cut the costs of unplanned downtime and emergency maintenance of their assets.

[1] VNI Global Fixed and Mobile Internet Traffic Forecast, Cisco