IM JAN 23 | Page 78

beginning of the shift , so fewer alarms will happen in the first place . Also , when an alarm does occur , they also want to be able to consult the operator ’ s ReadiScore at the time – essentially a second source of fatigue data to indicate which alarms are truly critical . This ultimately minimises disruption while maximising safety .”
In May 2021 , Hexagon ’ s Mining division announced that it was partnering with Fatigue Science to provide a complete , end-to-end visibility into operator fatigue , alertness and collision avoidance in a single offering . Hexagon is able to provide predictive fatigue insights by combining its MineProtect portfolio ( including its Operator Alertness System ) with actionable data from the Fatigue Science platform . Higdon said : “ We have expanded with Hexagon to a number of sites and our technical integration is getting stronger . The importance of integration has also been recognised . Now , a ReadiScore is displayed prominently as part of a Hexagon solution , superimposed over the camera feed . It ’ s a second source of fatigue data on the mine operator .”
He did add that there is still quite a lot of diversity in the fatigue system approach taken by mines – whether they want a camera based system first , or a combined system first as described , or a predictive solution first . “ You don ’ t need cameras to be able to provide the supervisor at the start of a shift with actionable information on which two or three workers who are most fatigued . In that sense , F / S is a strong independent value proposition .”
F / S has done studies comparing correlating microsleeps from camera systems to its ReadiScore , and have shown that where it predicts severe fatigue there is a much higher ( 14x ) likelihood of a video-verified microsleep occurring . “ That said , they are still different streams – our system is saying – at the start of a shift if and when you ’ ll have a high likelihood of a microsleep . In contrast , cameras detect microsleeps at the particular moment they occur .”
What types of action would the supervisor be willing to take at the start of a shift if they knew a particular individual was a much higher risk than others ? These actions include assigning extra breaks or re-assigning ultra-high-risk tasks before critical fatigue occurs . This proactive approach is often preferred stopping a truck midcycle in the event that a camera detects fatigue .
Finally , F / S since 2021 has offered ReadiSupervise Mobile , a new tool to help shift supervisors more effectively manage fatigue risk , improve operator safety , and increase crew performance in daily operations . It provides supervisors a complete workflow , allowing them to observe , predict , plan , and intervene on crew member fatigue – all from a simple mobile app . Supervisors can view ReadiScores , indicating fatigue level , for each crew member under their supervision , both in real-time and forecasted for each hour in the shift ahead . Already , the workflow element has been made much more streamlined for supervisors in terms of minimal numbers of taps they have to make to take actions . Another new element is a so-called ‘ hotspots ’ feature that reveals if that whole particular shift is more risky from a fatigue point of view for the whole crew , and at what time period . It is an aggregation of all operator ReadiScores .
Lastly , Higdon mentioned a market change in that many large mining companies and mines are releasing RFPs that specifically seek out predictive fatigue technology – this indicates it is being considered more at a corporate level than before . There has also been much more interest from mining contractors , and more recognition that having the technology in place can actually increase productivity rather than slow operations down . F / S recently released studies estimating an over 3 % productivity increase between a fatigued vs non-fatigued crew – and this helps the ROI case for the tech in the first place . IM
The power of voice
WOMBATT ’ s VOZ is a fatigue prediction technology based on analysis of the voice . Because WOMBATT uses artificial intelligence to analyse multiple voice muscles , accuracy once the algorithm learns the individual ’ s own voice characteristics is up to 90 %, higher it says than any other methods available . The company also says that as WOMBATT operates in the background , the system does not intrude on users and as such is popular with drivers and supervisors .
The technology and the algorithms behind it were developed by WOMBATT for the European Space Agency ( ESA ) Mars 500 simulated mission programme to predict and manage the fatigue of astronauts on long duration missions to the Moon , Mars and beyond . Besides mining it has already been deployed for operators in other industries including air traffic control and steel smelting amongst others – most recently it has begun a field trial with over 400 drivers with DHL in Oman which services the oil and gas sector in the country ’ s desert interior . Dan Wood , Managing Director , BE DHL Oman : “ In Oman , our drivers travel long distances through harsh environments on monotonous roads . As we continue our journey to ensure the safety of our drivers , it has been great to finally start a trial with the WOMBATT fatigue prediction technology with one of our cargo haulage contractors . The predictive technology will allow us to effectively avoid the risks of driver fatigue whilst planning our journeys . We look forward to concluding the trial and when successful , roll this out to all employees , not just drivers .”
WOMBATT VOZ analyses an 8 second voice audio recording , made by the operator using their mobile phone , tablet or computer and sent to an AI algorithm , which then takes approximately another five seconds depending on transmission speeds to predict the level of fatigue of the user and return the result to the user ’ s device , and in the event of an alert , SMS and / or emails to the shift supervisor , dispatch operator and control room . A green result is given as a " fit to work " reading for the next four to five hours . Orange means the person is at a higher risk of having a fatigue event but can continue work with higher awareness of their fatigue risk . Red means the risk is at point where action needs to be taken such as a power nap for the operator . The system requires only three recordings per day minimum , usually one at the start of the shift , one in the middle and one at the end . The first recording is mainly done in the line-up room where workers gather before the start of a shift . This is most practical where mines don ’ t allow mobile phones in cabs but tablets are also often used in this scenario . You can ’ t cheat and use a prerecording as every voice recording is unique . WOMBATT offers field trials free of charge and once installed uses a subscription-based model .
One of the company ’ s mining clients is currently a large gold mine in Alaska , which is a full commercial WOMBATT VOZ contract already two years in across the whole haul truck fleet ( over 50 Cat 793s ) as well as over 20 contractor haul trucks . This mine is also where the original commercial field tests of the technology were made that led to ESA approving its commercial release . In 2020 , the mine had 12 % WOMBATT red fatigue alerts out of all recordings ; in 2021 this had dropped to 5 % and for 2022 is around 2.5 %. This reflects the feedback loop – WOMBATT VOZ specifically looks for those with a lack of sleep – and predicts the likelihood of a microsleep so early preventative action can be taken . This then leads to the workers increasing their sleep time and quality , leading in turn to lower alerts . At the Alaskan gold mine , WOMBATT-VOZ runs on the Wenco mine dispatch system into which it is fully integrated . But it can run on any dispatch or journey management system , and equally the voice recording can be transmitted via an App to the Cloud or a local server at any time of the day or night by any Android , IOS or Windows 10 mobile phone , tablet or computer . Immediately when there ’ s an alert the driver and dispatch know simultaneously , which shows up on the dispatch or control room screen as well when integrated into the dispatch or journey management system , plus a text and email are sent to the supervisor . In 2023 , WOMBATT is set to also begin a trial with one of the major underground zinc miners in Tennessee .
74 International Mining | JANUARY 2023