Eyes on Air

Abstract We at ADNOC Logistics & Services have identified the need for a Fully Integrated Inspection and Monitoring Solution to meet our operational, safety and security objectives. It also helped us in our journey toward becoming a world-class integrated logistics services provider. We have a mandate to manage complex logistics operations while being flexible in services delivery by adopting the latest technology and leveraging strategic partnerships. ADNOC L&S adopted autonomous drone technology from Percepto in most of its critical operations. The artificial Intelligence in the drones automatically detects abnormal changes in working environment as well as unsafe acts and conditions and helps employees be more aware of them especially during routine activities. Finally, it helps management take immediate action to address unsafe conditions as soon as they occur. Visual inspections play a big role today in asset management. In fact, they're considered a best practice for ensuring safer and more productive operations. Being able to conduct visual inspections routinely leads to early detection of issues and damages that might become failures. In this way, visual inspections ultimately help minimize incidents. Yet visual inspections are not limited to preventing and minimizing incidents, but organizations also get value from real time monitoring of procedures such as planned shutdowns of specific assets such as a flare stack inspection. During construction, having the ability to monitor work that is being conducted in real time helps minimize the overall downtime. This can translate into saving hundreds of thousands and even millions of dollars. Inspections are vital and even crucial for business continuity. Yet, today visual inspection is far from being optimized. The end to end process is not at all efficient. And surprisingly, most companies and most sites still conduct visual inspections manually, not automatically. This type of inspection is labor intensive, takes a lot of time, and can even put employees at risk. Overall, manual inspection is an inefficient process. Consider the siloed workflows that comprise the overall inspection. You start with having to fetch the data and collect it. This involves sending people to the site with special equipment. It can also involve climbing up high structures or putting people into potentially dangerous positions. All of this is manual time-consuming work. When this is done, the data needs to be somehow transferred to people who are going to be analyzing it. They need to have a particular type of expertise and experience in managing visual data. Once they go over the data, they need to create or define some insights and share their findings with the relevant stakeholders. Yet again, this is a labor intensive and lengthy process. It's also costly. Fortunately, it does not have to work this way, as there is lots of room for automation. Each of the siloed workflows from autonomous capture visual data management to insights are starting to be automated. Note that typically, this is not an end to end automation. Organizations that are doing data capture of their facilities with drones, robots, piloted drones, are in most cases, automating data capture and are able to collect data more quickly, more easily without having to risk anyone. They are on the first step towards automation. Some organizations inspect their facilities with third party software that is capable of getting the data, storing it, and managing it in one place. This leaves them dependent on outside sources, which can slow the process and potentially be unreliable. Finally, there's some software solutions available that are unable to automate some of the insights. So, organizations have multiple options to begin automating inspections. Yet, none of these options provide a full holistic solution for inspection. This is where this technology comes into the picture with the AIM. AIM is the only end to end autonomous inspection and monitoring solution. It automates every step of the process holistically; from capturing the data, through managing it, all the way to sharing insights with different stakeholders. Let's look at the details related to each step of the process. First, there's automating data collection. We use autonomous robots in the form of a drone in a box. Companies have the autonomous drone on site, and they are fully operated by software. They are available as needed for running inspections. Companies can schedule inspections when they want and as often as they want – without outside assistance. This is the first step towards holistic automation. The second aspect of holistic automation has to do with the visual data management. It's not enough to automate the data collection, which is very complex in and of itself. Of what value is the data if it is not able to be used? Therefore, it's even more important to be able to gather all the data, store it in one place, tag it, and sift through it in order to eventually gain insights. And it should be easy to share those insights with relevant stakeholders from various departments. This promotes teamwork and allows everyone to feel involved in the process and add their own valuable insights. A user can simply request information for specific assets, and drones will be dispatched automatically. Plus, insights will be delivered for the user or users who requested the data. While this may sound futuristic, we at ADNOC are already using it to inspects our sites. The concept behind the operation of the AIM is fairly simple. It starts with deploying autonomous drones on site. Companies gain high quality insights with once-a-day inspection. The frequency of inspections and their consistency, because it's the same robots that go day after day and collect the data, lead to high level insights. The same robot or drone collecting data at the same time each day allows for regular comparison and strengthens anomaly detection. This is exactly what companies need - get the right insights that help ensure asset integrity, the completion of proper preventive maintenance, etc. A few concerns arise however when considering adopting an autonomous drone solution: namely - drone regulation and privacy. Drone regulators are understandably concerned about drones flying over people, operating remotely and in darkness or inclement weather, and – critically – the dangers of uncoordinated flights interfering with other aircraft. That's why when going through all the necessary requirements with all authorities in getting the approvals for autonomous to gained approval to operate drones beyond visual line of sight (BVLOS), which enabling this solution to be operated remotely without the need for an operator on site. When talking about drones and privacy, there's a huge difference between piloted recreational and autonomous industrial drones. Recreational drones cause the majority of drone accidents because they're flown by unlicensed pilots in public spaces with minimal supervision. Autonomous industrial drones fly in remote or rural locations, far from people, and have multiple built-in failsafes and advanced safety features like parachutes to eliminate injuries and mishaps. Now - let's take a deep dive to learn how drones in autonomous inspection solutions can benefit refinery inspections. We've been deploying the AIM for their end to end solution as part of our vision of creating and being part of the future. It is integrating multiple cutting edge technologies in order to guarantee safer and more productive operations. Different teams – from operations to maintenance, to emergency response and others - in our organization are getting value from having this technology. Note, the technology is applicable to any type of industrial site, construction site, and of course, oil and gas industry. Regardless, they all benefit from the autonomous inspection and monitoring technology to cater different use cases across the lifecycle, from construction of a new site, through the ongoing operation, we're using it for emergency response, and even for periodic complex projects, such as turnarounds. Companies have deployed the solution to monitor the construction of new refineries that they are building. Having the autonomous robots on site daily is especially helpful because it enables them to collect data and easily monitor what's going on throughout the construction work. It enables companies to ensure that schedules are met, verify that the work has been completed, and that it is consistent with the plans. In one instance, a company was able to recognize that contractors were not fulfilling their promise. The contractors were supposed to send a team that included a specific number of people. By having a drone provide visibility of what was going on at the site with this the supervisors were able to count how many people were on site. The supervisors realized that the actual size of the team on site was much smaller than promised. By being aware of this sort of thing and being able to verify it, whether in real time or at the end of the day, a company can make better decisions and be clear if things are proceeding according to plan. Having the drones on site collecting visual data and providing companies with visibility to know what's going on throughout the construction work. It can also help to enforce compliance with both environmental and safety standards. Consider, the drone is streaming data in real time, and generating reports at the end of the day, and/or at the end of the week, depending on the configuration done on site. The reports can be checked for things such as whether people were wearing their safety gear, their PPE, whether they were going on to restricted areas. In addition to allowing for awareness, it also it's a form of proof as to what happened on site that you can use later, should the need arise. Once construction is complete, it's time to consider ongoing operations. This is where the largest number of use cases can be found in the day to day operations of a refinery or any other oil and gas facility. An autonomous inspection and monitoring solution is used for various use cases, including inspections of tanks and pipes, and chimneys and flare stacks. It can be using a high-res visual camera for inspection or thermal inspection. Since the drones are also equipped with onboard artificial intelligence capabilities, AIM is also capable of detecting things like oil leaks, gas leaks, structure failures and more. One of the great advantages of having such high frequency inspections with the very same drones collecting data day after day or week after week is the fact that it leaves a company with historical data. And when applying AI, it can provide a lot of insights that otherwise (think periodic inspections or even manual inspections which capture a fraction of what is going on) would have been very difficult to gain. Here are some other examples of how using the historical data and applying anomaly detection is helpful. It can easily detect things like rust on top of tanks, or objects that were left behind. These things start off as minor things but if left to grow can become significant problems. By the way, the drones are equipped with thermal cameras to see things that are not even visible at all, when you're just flying piloted drones. We found this helpful when on one of the very first flights, we realized with thermal imagery that we had a problem with insulation of pipe that was going on underground. Furthermore, the solution enabled to have a very early detection of this problem. And this again only happens because of high frequency inspections, and also visual and thermal and applying AI at the same time to gain invaluable insights and to be alerted on time. Again, a key element to keep in mind regarding the solution is that a person doesn't have to be there in front of the system analyzing what the robots are collecting. This is done via AI which is able to provide valuable insights. Next, there's emergency response as it is related to both security and safety. First of all, the solution can be used to augment the security teams and run security patrols. This can be done with drones. Companies can schedule the system to go on as many patrols a day that is appropriate for their facility. These patrols augment the security team and provide a lot more situational awareness. Another security benefit is being able to apply AI in the form of detection and tracking of people and vehicles. So, as soon as the robots identify suspicious objects, they can start to automatically track that object. People can make decisions about the necessity to take action or not from a safe distance. Companies also can control the robots ad hoc for immediate response. In the control room, companies get a call saying that people identified smoke. Upon hearing this, a drone is dispatched. Within minutes, they have situational awareness, even before the first responders even got on to the relevant area. This enables companies to direct the first responders as to where the smoke is coming from and how to get to it, which alleviates the risk they would have experienced if they went in blind. Another example is being able to do advance planning for storms and excessive temperatures. Now, you can never be completely prepared for an emergency. But with the solution available on hand, preparation can occur before the storm occurs in the form of running a quick inspection to see whether some areas are more prone to failure and issues could arise in that location due to the storm or excessive temperatures. Once the storm is over, the drones can be quickly dispatched to identify those areas that became faulty and require maintenance. This saves valuable time as instead of sending maintenance crews to go and look for problems, maintenance crews can be immediately dispatched to fix the problems that were spotted by the drones. Keeping service in operation and minimizing downtime is sure to please customers and clients. It can also mean a lessening of environmental damage. The last major lifecycle phase, where the solution is being used, is turnaround. In terms of turnaround, suddenly there are hundreds of employees that are not necessarily familiar with the site going there for the sake of conducting the turnaround work. Remember, every day and even every hour and minute beyond the expected turnaround time means a loss of money for a facility. Autonomous inspection and monitoring can help minimize this in three ways. Time and tools help with productivity. During turnaround times, often there are long lines that form – it could be during lunch, break times, upon arrival in the morning, a changing of the shift in the morning, etc. During these times, when employees are getting in and out of the facility, the robots can provide visibility and maintain visual awareness. Another benefit is that by knowing where lines are building up, a remedy might be found that can mitigate the issue and get people back to work quicker so more time is spent on the job. Maximizing time and tools can translate into a lot of money for this, you know, throughout these times of turnaround. Another element is safety. The drones can stroll around the facility to look for safety hazards. Perhaps, an employee accidentally left equipment behind, or service materials were not correctly left, etc. By monitoring the grounds, potential safety hazards can be reported in real time to supervisors in the control room, allowing them to ensure that safety is maximized for everyone on site. A final element related to turnaround time is that of the washout areas. Being able to tag the different components that are getting in and out and say exactly when they got in and where they were left once, they were washed properly, etc. - again, guaranteeing that plant work is being conducted as planned and nothing is getting lost. A logical question to ask – is it worth it? Well, there are two main drivers behind ROI and the first one is reduced costs. Being able to automate the labor intensive, lengthy manual inspection, translates directly into savings. The costs involved with manual inspection from insurance to special equipment to time spent by employees for doing the work can run hundreds of thousands of dollars. Of course, most companies inspect their facilities a few times a year, leading to an exorbitant amount of money needed for inspections. The second significant contributor to ROI is increased productivity/increased safety. By minimizing incidents and maximizing uptime, ensuring business continuity, companies gain greatly. Keeping employees safe helps to boost morale and improve production. This technology is truly capable of automating everything end to end and has proven its value. We at ADNOC have found the main benefit is simplifying the detection and control of unsafe acts and unsafe conditions in a dynamic and high risk work environment such as the maritime industry. Percepto helps us improve our company employees’ ability to remain safe, which in turn leads to a reduction in downtime and insurance costs. All this further enhances our reputation and protects ADNOC's brand name. It demonstrates collaboration, as well as motivation and encouragement of others to enhance or resolve occupational HSE issues in exhibiting the will and ability to intervene when required to stop unsafe actions and conditions.