The Use of IoT to Earlier Identify Faults on Lifts

Paul Clements, Associate at D2E International VT Consultants Ltd, recently graduated with an MSc in Lift Engineering from the University of Northampton during which he submitted his dissertation entitled; “The use of IoT technology to enable earlier identification and monitoring of faults on lifts”.

Below we provide a review of Paul's dissertation paper, which you can read in full here.

His logical approach to this vast subject has seen Paul summarise the history of IoT from the early days of ‘smart devices’ in 1982 through to the latest developments which are bringing IoT devices to the market in thousands of different and affordable formats. Paul has taken his understanding of IoT opportunities and coupled this with his in-depth knowledge of the Vertical Transport (VT) market and more specifically his understanding of how lift control systems work in our modern world.

Paul highlights how modern microprocessor-based control systems enable intelligent fault-finding based upon detection, localisation, assessment and more recently prediction. Major lift companies; Otis, Kone, Schindler and Thyssenkrupp are now focusing huge research and development budgets into predictive, preventative and even pre‐emptive maintenance based around the use (in various forms) of IoT.

Trawling through relevant VT codes (BS EN81-20:2014), Paul has linked published parameters to potential measurements which could be gathered using IoT sensors such as; machine temperature, motor room conditions, lift car levelling, noise restrictions, vibration tolerances etc. Paul has also reviewed best practice guidance (CIBSE Guide D) reviewing other measurable parameters such as ride comfort, noise, vibrations and humidity.

IoT in the VT industry is not a completely new concept and Paul has called upon his experience at Blackpool Towers where IoT sensors have enabled close monitoring of wind speeds to ensure the safe use of the tower passenger lifts in windy weather conditions. This example demonstrates how IoT devices can collect data and detect errors; the analysis of this ‘Big Data’ is critical in delivering benefits to suppliers, users and clients.

Paul gathered a large amount of data (over 800 faults) from various sources over a period of 13 months to ascertain the key faults with lifts. This analysis has been used by Paul to establish which lift components IoT sensors should measure. The analysis found that the key reasons for callouts are; No Fault Found (NFF)/Running on Arrival (ROA) (18%), Controller Reset (13%), Door Lock Adjustments (8%), Misuse (6%) and Audodialer faults (6%). When grouped into component related areas the split is; Doors (25%), Motor Room (20%), Lift Car/Landing (20%), Shaft (10%) and Other (25%).

Paul selected a simple, cost-effective ‘IoT sensor’ option which could measure; light, temperature, movement, humidity, open/close and moisture. Paul identified that using these ‘IoT sensors’ 20 of the key faults identified could be monitored by this technology covering some 45% of all faults along with the majority of the NFF/ROA faults which account for an additional 18% of the faults.

Paul extended his research by questioning industry experts into common fault types, areas of focus and what can/can’t be measured using IoT sensors. These industry experts were in agreement with Paul’s VT fault analysis which gave him the confidence to put his theory into practice by moving his research into live experiments.

Paul purchased a series of sensors, routers and associated equipment and then went through a frustrating period of overcoming substantial hurdles such as; equipment not functioning, interface issues, firewalls, the transmission of data, battery power etc… Sensibly, Paul decided to start his experiment in a ‘safe’ environment by setting up a basic monitoring system at home before transferring this experience to a live lift site. After several failed attempts a working prototype was finally set up and the data started to flood in.

A sensor was installed to measure the movement of the main sheave which gave live data to indicate when the lift was moving or stationary. An alert was set up if the sheave was inactive for one hour indicating a possible failure in the lift. This data enabled a detailed analysis of the lift usage pattern over a full week identifying peak times and non-active times. The same sensor measured the temperature of the Motor and any humidity around the motor. This enabled Paul to establish the normal temperature and humidity range (upper and lower limits) with alerts if these were broken. An additional sensor in the machine room measured the temperature within the machine room to ensure it was within parameters set under code.

Following the success of this trial, Paul moved to a new site and tested new sensors positioned in new places (including the pit) and also introduced the use of a camera to view the fault logger. During this trial, Paul received a notification that the lift hadn’t moved for an hour, he checked the fault logger via the camera and identified a fault with the lift. He notified the maintenance company who attended site and rectified the issue before the client was even aware there was an issue! Success! As a final test, Paul also attempted successfully to log onto the system from various locations around the world to prove the system is truly portable.

Paul has learnt a great deal from these trials in terms of;  where to locate the equipment, the need for constant power supply, battery life of sensors, what works and what doesn’t. Ingenuity and thinking outside the box has taught Paul that there are many ways to skin a cat; the movement of lift can be monitored by a movement sensor on the main sheave or by vibration on the lift motor.

Sensors and IoT technology is a cheap commodity product in use in everyday life for thousands of different applications, the VT industry is playing catch up. Major lift companies are investing heavily in this technology but still seem to be inward-looking with the ‘what can this do for me mentality’. Paul has proven in his dissertation that IoT’s can; effectively monitor key components within a lift system, detect faults and effectively communicate these to maintenance providers, users etc. He has also demonstrated that this can be done very cheaply using off-the-shelf devices.

The technology works but so what? Paul completes his extensive study by looking at future developments and alternative uses of this technology; escalators and façade access equipment and the use of different sensors to measure different elements. The sky is the limit. The technology exists and has now been proven but how is this taken to the market? What are the benefits to customers? Will this reduce maintenance costs? Will this improve the uptime of equipment? Could we measure ride quality? How about lift overcrowding?

If you would like to learn more about how lifts are playing a vital role in Smart Buildings and IoT technology is changing the lift industry, please feel free to contact Paul on +44 [0] 7825 505 994 or email paul.clements@d2e.com.

- Review by Mark Fairweather. The full paper can be read here.