Predictive Maintenance – A Comparative Study on Startups

Posted by Machine Sense on

A new industrial revolution is here and it’s called Industry 4.0! It has transformed the way machines collect and interpret data. The Industrial Internet of Things (IIoT) plays a pivotal role in this process by increasing the efficiency of predictive maintenance. This phenomenon is changing the manufacturing industry from its core. The market is projected to grow at a whopping 22.1% from USD 4.56 billion in 2017 to USD 12.41 billion by 2022.

In the manufacturing industry, ‘smart factories’, created through Industry 4.0 are making a seamless transition from the traditional approach to the Artificial Intelligence (AI) approach. As a result of the emerging trend, several disruptive technology companies have made the market for predictive maintenance highly competitive. These companies are using machine sensors to collect data and the predictive analytics is decreasing downtime up to 45%. Furthermore, this is increasing productivity by 25%. It is the age of creating new business models through proactive monitoring which results in better customer service.

A good example is Israeli startup Augury. Their product ‘listens’ to machines in order to do predictive maintenance. They claim that their product results in 70% fewer breakdowns through preventive and predictive maintenance. Augury's platform automatically diagnoses machines based on the sounds that they make. This is achieved by connecting vibration and ultrasonic sensors to smartphones and pairing them with machine learning algorithms. Securing USD 17 Million in Series B funding, Augury is considered to be one of the leading providers of machine diagnostics through Industrial Internet of Things.

Another established start-up, C3IoT, offer their platform as a service that rapidly develops and deploys AI based solutions. Raising another USD 100 million from its existing investors, C3 IoT is gaining momentum providing predictive insights for cross-industry business challenges. With the US Department of Defense Innovative Unit and 3M as their clients, C3IoT is certainly considered a world leader in IoT and AI applications. Microsoft and C3IoT announced a strategic partnership to create solutions using the C3IoT Platform which eventually will integrate and operate on the Microsoft Azure Cloud.

The most promising of all probably is Baltimore-based startup MachineSense. With R & D facilities in India, they have a solid manufacturing background and that augmented with tech savvy coders make them formidable in this space. Their patented solutions allow their clients to predict malfunction and reduce downtime using electricity and vibration analytics. Recognizing their potential, Microsoft Corp has chosen MachineSense to be part of their elite ScaleUp accelerator program in 2018. Among many other USPs, they are the only company in this space who have the ability to run analytics fully, both at the edge and in the cloud.

Not to be ignored also is Uptake. Founded in 2014, their machine learning algorithms analyze data from equipment sensors, learning the normal operation patterns and identifying upcoming failures. They raised USD 117 million, taking their total funding to USD 250 million. They describe themselves as one of the largest SaaS players offering a plug and play solution to their clients.

And finally there is Petasense. Based in Silicon Valley and with seed funding of USD 1.8 million, they offer end-to-end solutions that include wireless vibration sensors, cloud software and machine learning analytics. In the Petasense model, asset reliability and predictive maintenance improve through data gathered over time because false positives and false negatives are greatly reduced. However, Petasense is currently providing solutions for rotating machines only.

The key parameters that allow the above companies, active in the factory maintenance market, to achieve a strategic advantage are:

  1. Sensor Diversity: The predictive maintenance system depends on a myriad of sensor data for monitoring the condition of a machine. They include vibration, current, voltage, temperature, magnetic field, rotation speeds, etc. It is imperative that companies focus on collecting data from these diversified channels to get comprehensive information. MachineSense encompasses the entire gamut when compared to companies like Augury, C3IoT who focus on vibration and sound or even 3rd party sensors.
  2. Stream Analytics: Using a combination of Time Domain analysis and Frequency Domain (FFT) analysis will result in identifying abnormal deviations as well as the diagnosis of specific faults. MachineSense has a rich statistical ensemble extracted from both the analyses. Most start-ups like Petasense use FFT or 3rd party streaming only.
  3. Actionable Vertical Analysis: This forms the fundamental part of predictive maintenance. The analysis can identify early signals of disastrous failure that can avoid unplanned downtime. The manufacturer can improve the designs leading to optimal operations and cost savings. Prescriptive maintenance analysis will allow the manufacturer to know, e.g., when to change a filter, oil, etc. Actionable Vertical Analysis also examines machine utilization and verifies the real operating cost and energy efficiency. Only MachineSense, Uptake and C3IoT take into account Machine Utilization and Energy and Cost analytics.
  4. Distributed Analytics: Augury is ahead from its competitors in this regard as it uses Historical Analytics to ascertain how both machines and processes operate. MachineSense closely follows with quasi real time, real time and 24 x 7 analytics. The data allow the end users reliable actionable intelligence and better security.
  5. Open Data Architecture: Companies looking for alliances and expansions have to rely on open data architecture for its high end-user impact. From predictive and preventative maintenance to prescriptive recommendation, MachineSense offers everything when compared to Augury that does not offer the solution. C3IoT, Petasense and Uptake partially include open data architecture for their clients.

Smart factories are evolving everyday with new AI technology improving how machines talk to machines. In this high-stakes market, Microsoft Corp. is heavily investing in companies that are pioneers. MachineSense and C3IoT are the very few that are working along with Microsoft on their Azure platform. The association will help these startups with go-to-market strategies, technology and community building.

The future of the factory maintenance market is robust with data driven decisions. As an emerging market, machine learning technologies will progressively enhance how companies plan and operate their business.

Older Post Newer Post

0 comments

Leave a Comment

Please note, comments must be approved before they are published