Start Condition Monitoring With Vibration Sensors

IIOT (Industrial internet of things) has gained much traction and attraction in past years. With industries getting their assets online for monitoring purposes and new IO-Link sensors providing a ton of information on a single package, monitoring machines has become economically feasible.

Vibration is one of the most critical metrics regarding the health of machines, providing early detection of potential faults – before they cause damage or equipment failure. But since this is a relatively new field and use case, there is not much information about it. Most customers are confused about where to start. They want a baseline to begin monitoring machines and then finetune them to their use case.

“Vibration is one of the most critical metrics regarding the health of machines…”

One approach to solve this is to hire a vibration expert to determine the baseline and the best location to mount the vibration measuring sensor. Proper setup increases the threshold of getting into condition monitoring as a new user figures out the feasibility of such systems.

I direct my customers to this standardized baseline chart from ISO, so they can determine their own baselines and the best mounting positions for their sensors. The table shows the different standards for severity for different machine classes. These standards detail the baseline vibration and show the best place to mount the sensor based on the machine type.

Click here for more information on the benefits of condition monitoring.

 

Manufacturing Insights: Top Blogs From 2021

While last year was filled with challenges and unexpected changes for many industries, including manufacturing, it was not without positive achievements and insights. As we look forward to 2022, let’s not forget some topics that shaped 2021, including our five most-read blogs.

1. 5 Manufacturing Trends to Consider as You Plan for 2022

 

 

 

 

It’s that time of year again where we all start to forget the current year (maybe that’s OK) and start thinking of plans for the next – strategy and budget season! 2022 is only a few weeks away! I thought I’d share 5 insights I’ve had about 2022 that you might benefit from as you start planning for next year.

READ MORE>>

2. The Pros and Cons of Flush, Non-Flush and Semi-Flush Mounting


Inductive proximity sensors have been around for decades and have proven to be a groundbreaking invention for the world of automation. This type of technology detects the presence or absence of ferrous objects using electromagnetic fields. Manufacturers typically select which inductive sensor to use in their application based on their form factor and switching distance. Although, another important factor to consider is how the sensor will be mounted.

READ MORE>>

3. IO-Link Wireless – IO-Link with Even Greater Flexibility



In a previous blog entry, I discussed IO-Link SPE (Single-Pair Ethernet). SPE, in my opinion, has two great strengths compared to standard IO-Link: cable length and speed. With cable lengths of up to 100 meters and speed of 10 Mbps, compared to 20 meters and max baud rate of 230.4 Kbps, what could be out of reach?

READ MORE>>

4. How Condition Monitoring has Evolved and Its Role in IIoT

In recent years, as IIoT and Industry 4.0 have become part of our everyday vocabulary, we’ve also started hearing more about condition monitoring, predictive maintenance (PdM) and predictive analytics. Sometimes, we use these terms interchangeably as well. Strictly speaking, condition monitoring is a root that enables both predictive maintenance and predictive analytics. In today’s blog we will brush up a little on condition monitoring and explore its lineage.

READ MORE>>

5. Lithium Ion Battery Manufacturing – RFID is on a Roll



With more and more consumers setting their sights on ‘Drive Electric,’ manufacturers must prepare themselves for alternative solutions to combustion engines. This change will no doubt require an alternative automation strategy for our electric futures.

READ MORE>>

Honorable Mention: Top 5 Insights From 2020

And, finally, for the sake of comparison, we can’t help but honorably mention last year’s look-back blog. The top five insights from 2020 include buying a machine vision system; data provided by IO-Link; changes in electrostatic sensing field by capacitive sensors; reducing the number of ethernet nodes on your network using IO-Link; and adding a higher level of visibility to older automation machines.

Read more>>

We appreciate your dedication to Automation Insights in 2021 and look forward to growth and innovation in 2022!

3 Easy Options to Get Started With IIoT in 2022

The Industrial Internet of Things (IIoT) may seem large, intimidating, and challenging to implement; however, new systems and solutions will eliminate the perceived barriers for entry. As we wrap up the year and make plans for 2022, now is a great time to resolve to modernize your facility.

Do you have a process, system or machine that has outlived its life expectancy for many years or even decades and isn’t up to current IIoT standards? Great news: you have several options for updating.

Traditional approach

The traditional approach allows you to use your current controller to output your information to your existing database. If you want to try IIoT on your current setup and your controller cannot be modified, a self-contained system will allow for ultimate flexibility. It will provide you with access to the data based off an extra layer of sensing with a focus on condition monitoring. This approach is the least expensive route, however, if database access is restricted the following options may be better choices.

Cloud-based current industry standard

A second option is to use a portable monitoring system that has a condition monitoring sensor. It is essentially five sensors in one package that can hook up to a system using the cellular network to report data to a secure cloud database. This approach is useful in remote locations or where local network access is limited. If you have a problem area, you can apply this temporarily to collect enough data, enabling you to implement predictive maintenance.

Local-based current industry standard

A local self-contained system is a great solution if a cloud database is not desired or allowed. Systems such as a Condition Monitoring Toolkit allow for recording of devices onto the local memory or USB drive. Additionally, multiple alarm set points can be emailed or extracted locally. This approach is best for testing existing machines to help with predictive maintenance, to improve a process, or even to prevent a failure.

All three of these options require data management and analysis to improve your processor and to remedy problematic areas. Using any of them is an opportunity to test the IIoT waters before fully diving in. Extrapolating the results into problem-solving solutions can allow you to expand IIoT to the rest of your facilities in a cost-effective manner.

Automation is “Rolling Out” in the Tire Industry

Automation is everywhere in a tire plant – from the old manual plants and mid-hybrid automated plants to the newest plants with the latest automation technology all over the world.

Industry challenges

Some tire industry automation challenges are opportunities for automation suppliers and machine builders. These can vary from retrofitting old machines and designing new machines to including smarter components to bring their production into the IIoT.

Plants want to save CapX dollars on new machines, so they are looking to upgrade old ones. Tire plants are learning from the past. They are limited by their older technology, but it has been hard to upgrade and integrate new technology, so there are long-term needs for adding flexible automation on machines. This requires new processes and recommissioning machines quickly. A good example of this is the addition of a vision system to improve quality inspections.

More automation is also needed due to a lack of skilled labor in the industry combined with the desire for higher throughout. The addition of robots on the line can aid with this. Plants can also simplify their wiring by migrating away from control panel i/o/analog to an IP67 network and IO-Link master and hubs.

The use of IO-Link also allows for more continuous condition monitoring. There is an increased need for quality inspections and process improvements. Plants are collecting more data and learning how to use it and analytics (Industry 4.0, IIoT) to achieve operational excellence. Plants need more technology that supports preventive and predictive failure solutions.

Additionally, there are automation needs on new machinery as tire designs are in an evolutional growth/change period – in the electric vehicle (EV) market, for example, where rapid change is happening across all vehicle manufacturing. Smart tires are being designed using RFID and sensors embedded in the tire ply.

Successfully matching up automation products to meet plant needs first requires understanding the plant’s main processes, each with millions of dollars of automation needs.

How tires are made

    1. Raw materials logistics – raw materials are transported to the mixing and extrusion areas for processing.
    2. Mixing and extrusions – up to 30 ingredients are mixed together for a rubber blend tire.
    3. Tire components – extruded rubber ply is measured and cut to size to meet the needs of the specific tire and then loaded onto reels feeding the tire building machines.
    4. Tire build machines – tires are built in stages from the inside out. They are crated without tread and transferred to the curing press machines.
    5. Tire curing press machines – here, the “green” tires are vulcanized, a chemical process that makes the tire more durable. Tire parts are then compressed together into the final shape and tread pattern.
    6. Inspection and test machines – tires are quality tested and undergo visual, balance, force, and X-ray inspections.
    7. Logistics material handling, conveyor, ASRS, AGV – finished tires are taken to the warehouse for sorting and shipping.

In the past, not many people outside the tire industry understood the complexity and automation needs of these high volume, high quality, highly technical plants. Tires are so valuable to the safety of people using them that manufacturers must be held to the highest standards of quality. Automation and data collection help ensure this.

In the meantime, check out these futuristic tires and imagine all the automation to manufacture them.

Controls Architectures Enable Condition Monitoring Throughout the Production Floor

In a previous blog post we covered some basics about condition monitoring and the capability of smart IO-Link end-devices to provide details about the health of the system. For example, a change in vibration level could mean a failure is near.

This post will detail three different architecture choices that enable condition monitoring to add efficiency to machines, processes, and systems: in-process, stand-alone, and hybrid models.

IO-Link is the technology that enables all three of these architectures. As a quick introduction, IO-Link is a data communications technology at the device level, instead of a traditional signal communication. Because it communicates using data instead of signals, it provides richer details from sensors and other end devices. (For more on IO-Link, search the blog.)

In-process condition monitoring architecture

In some systems, the PLC or machine controller is the central unit for processing data from all of the devices associated with the machine or system, synthesizing the data with the context, and then communicating information to higher-level systems, such as SCADA systems.

The data collected from devices is used primarily for controls purposes and secondarily to collect contextual information about the health of the system/machine and of the process. For example, on an assembly line, an IO-Link photo-eye sensor provides parts presence detection for process control, as well as vibration and inclination change detection information for condition monitoring.

With an in-process architecture, you can add dedicated condition monitoring sensors. For example, a vibration sensor or pressure sensor that does not have any bearings on the process can be connected and made part of the same architecture.

The advantage of an in-process architecture for condition monitoring is that both pieces of information (process information and condition monitoring information) can be collected at the same time and conveyed through a uniform messaging schema to higher-level SCADA systems to keep temporal data together. If properly stored, this information could be used later for machine improvements or machine learning purposes.

There are two key disadvantages with this type of architecture.

First, you can’t easily scale this system up. To add additional sensors for condition monitoring, you also need to alter and validate the machine controller program to incorporate changes in the controls architecture. This programming could become time consuming and costly due to the downtime related to the upgrades.

Second, machine controllers or PLCs are primarily designed for the purposes of machine control. Burdening these devices with data collection and dissemination could increase overall cost of the machine/system. If you are working with machine builders, you would need to validate their ability to offer systems that are capable of communicating with higher-level systems and Information Technology systems.

Stand-alone condition monitoring architecture

Stand-alone architectures, also known as add-on systems for condition monitoring, do not require a controller. In their simplest form, an IO-Link master, power supply, and appropriate condition monitoring sensors are all that you need. This approach is most prevalent at manufacturing plants that do not want to disturb the existing controls systems but want to add the ability to monitor key system parameters. To collect data, this architecture relies on Edge gateways, local storage, or remote (cloud) storage systems.

 

 

 

 

 

 

The biggest advantage of this system is that it is separate from the controls system and is scalable and modular, so it is not confined by the capabilities of the PLC or the machine controller.

This architecture uses industrial-grade gateways to interface directly with information technology systems. As needs differ from machine to machine and from company to company as to what rate to collect the data, where to store the data, and when to issue alerts, the biggest challenge is to find the right partner who can integrate IT/OT systems. They also need to maintain your IT data-handling policies.

This stand-alone approach allows you to create various dashboards and alerting mechanisms that offer flexibility and increased productivity. For example, based on certain configurable conditions, the system can send email or text messages to defined groups, such as maintenance or line supervisors. You can set up priorities and manage severities, using concise, modular dashboards to give you visibility of the entire plant. Scaling up the system by adding gateways and sensors, if it is designed properly, could be easy to do.

Since this architecture is independent of the machine controls, and typically not all machines in the plant come from the same machine builders, this architecture allows you to collect uniform condition monitoring data from various systems throughout the plant. This is the main reason that stand-alone architecture is more sought after than in-process architecture.

It is important to mention here that not all of the IO-Link gateways (masters) available in the market are capable of communicating directly with the higher-level IT system.

Hybrid architectures for condition monitoring

As the name suggests, this approach offers a combination of in-process and stand-alone approaches. It uses IO-Link gateways in the PLC or machine controller-based controls architecture to communicate directly with higher-level systems to collect data for condition monitoring. Again, as in stand-alone systems, not all IO-Link gateways are capable of communicating directly with higher-level systems for data collection.

The biggest advantage of this system is that it does not burden PLCs or machine controllers with data collection. It creates a parallel path for health monitoring while devices are being used for process control. This could help you avoid duplication of devices.

When the devices are used in the controls loop for machine control, scalability is limited. By specifying IO-Link gateways and devices that can support higher-level communication abilities, you can add out-of-process condition monitoring and achieve uniformity in data collection throughout the plant even though the machines are from various machine builders.

Overall, no matter what approach is the best fit for your situation, condition monitoring can provide many efficiencies in the plant.

5 Manufacturing Trends to Consider as You Plan for 2022

It’s that time of year again where we all start to forget the current year (maybe that’s OK) and start thinking of plans for the next — strategy and budget season! 2022 is only a few weeks away!

I thought I’d share 5 insights I’ve had about 2022 that you might benefit from as you start planning for next year.

    1. Electric Vehicles

      The electric vehicles manufacturing market is receiving major investments, machine builders are building up expertise, and consumers are trending towards more electric vehicles. According to PEW research, 7% of US adults say they currently own a hybrid or electric vehicle, but 39% say the next time they purchase a vehicle they are at least somewhat likely to seriously consider electric. Traditional automotive won’t go away any time soon, but I see this as a growth generator.

    1. Automation in Agriculture & Food

      Automation in the agriculture, food, beverage and packaging markets is also growing strong with more demand for packaged goods and more SKUs than ever before. Urbanization and shortages in agriculture labor markets are driving investments in automation technologies in manufacturing and on the farm. Robotic agriculture startups seem to be growing faster than weeds and are providing real value for those who are struggling to get product from the field to the factory.

    1. Supply Chain Disruption

      Several economists have said the chip shortage will be with us well into 2023, and now I hear rumors of plastics or other materials having disruptions. Disruption might be the new normal for the short to mid-term. I flew out of LAX a few weeks ago and there were dozens of container ships parked outside the port. We are also seeing a major breakdown of our “over-land” logistics infrastructure. Investment in automation and labor for this market will be vital to a strong recovery. Plan for these things and be willing to have open and honest discussions with your vendors and your customers. Untruths might get you by in the short term but could permanently damage your business relationships for years.

    1. Real not Hyped Sustainability

      As Generation Z (18-24year old) workers increasingly enter our economy, they are pushing us to truly work towards sustainability much more than Millennials did before them. What this means is other markets that I see as growth opportunities are ones where we can have major impact on this, like mining, waste/recycling, and agriculture.

    1. Technology as an HR tool

      All manufacturers will be impacted by the skills-gap and labor shortage if you aren’t already. Part of your strategy for 2022 must include automation and robotics as part of your labor strategy. We need to consider how can we use automation and robotics to do our dull, dirty, dangerous jobs or how can we use automation and robotics to extend the careers of our long-term experienced workers. What disruptive technology could you be investing in to make a real difference in your work processes — 3D printing, machine vision, AR/VR, exoskeletons, drones, virtual twin, AI, predictive maintenance, condition monitoring, smart sensors? Pick something you will do different in 2022. You have to.

What do you see for 2022 that will have a major impact on our businesses?

How Condition Monitoring has Evolved and Its Role in IIoT

In recent years, as IIoT and Industry 4.0 have become part of our everyday vocabulary, we’ve also started hearing more about condition monitoring, predictive maintenance (PdM) and predictive analytics. Sometimes, we use these terms interchangeably as well. Strictly speaking, condition monitoring is a root that enables both predictive maintenance and predictive analytics. In today’s blog we will brush up a little on condition monitoring and explore its lineage.

Equipment failures have been around since the beginning of time. Over the years, through observation (collecting data) and brute-force methods, we learned that from time-to-time every piece of equipment needs some TLC. Out of this understanding, maintenance departments came to existence, and there we started having experts that could tell based on touch, smell and noise what is failing or what has gone wrong.

Figure 1: Automation Pyramid

Then we started automating the maintenance function either as a preventative measure (scheduled maintenance) or through some automated pieces of equipment that would collect data and provide alerts about a failure. We proudly call these SCADA systems – Supervisory Control and Data Acquisition. Of course, these systems did not necessarily prevent failures, but help curtail them.  If we look at the automation pyramid, the smart system at the bottom is a PLC and all the sensors are what we call “dumb sensors”. So, that means, whatever information the SCADA system gets would be filtered by the PLC. PLCs were/have been/ and are always focused on the process at hand; they are not data acquisition equipment. So, the data we receive in the SCADA system is only as good as the PLC can provide. That means the information is primarily about processes. So, the only alerts maintenance receives is when the equipment fails, and the process comes to a halt.

With the maintenance experts who could sense impending failures becoming mythological heroes, and  SCADA systems that cannot really tell us the story about the health of the machines, once again, we are looking at condition monitoring with a fresh set of eyes.

Sensors are at the grass root level in the automation pyramid, and until the arrival of IO-Link technology, these sensors were solely focused on their purpose of existence; object detection, or measurement of some kind. The only information one could gather from these sensors was ON/OFF or a signal of 4-20mA, 0-10V, and so on. Now, things are different, these sensors are now becoming pretty intelligent and they, like nosy neighbors, can collect more information about their own health and the environment. These intelligent sensors can utilize IO-Link as a communication to transfer all this information via a gateway module (generally known as IO-Link master) to whomever wants to listen.

Figure 2: IO-Link enabled Balluff photo-eye

The new generation of SCADA systems can now collect information not only from PLCs about the process health, but also from individual devices. For example, a photo-eye can measure the intensity of the reflected light and provide an alert if the intensity drops beyond a certain level, indicating a symptom of pending failure. Or a power supply inside the cabinet providing an alert to the supervisory control about adverse conditions due to increase temperature or humidity in the cabinet. These types of alerts about the symptoms help maintenance prevent unplanned downtime on the plant floor and make factories run more efficiently with reduced scrap, reduced down-time and reduced headaches.

Figure 3: The Next Generation Condition Monitoring

There are many different condition monitoring architectures that can be employed, and we will cover that in my next blog.

Improve OEE, Save Costs with Condition Monitoring Data

When it comes to IIOT (Industrial Internet of Things) and the fourth industrial revolution, data has become exponentially more important to the way we automate machines and processes within a production plant. There are many different types of data, with the most common being process data. Depending on the device or sensor, process data may be as simple as the status of discrete inputs or outputs but can be as complex as the data coming from radio frequency identification (RFID) data carriers (tags). Nevertheless, process data has been there since the beginning of the third industrial revolution and the beginning of the use of programmable logic controllers for machine or process control.

With new advances in technology, sensors used for machine control are becoming smarter, smaller, more capable, and more affordable. This enables manufacturers of those devices to include additional data essential for IIOT and Industry 4.0 applications. The latest type of data manufacturers are outputting from their devices is known as condition monitoring data.

Today, smart devices can replace an entire system by having all of the hardware necessary to collect and process data, thus outputting relative information directly to the PLC or machine controller needed to monitor the condition of assets without the use of specialized hardware and software, and eliminating the need for costly service contracts and being tied to one specific vendor.

A photo-electric laser distance sensor with condition monitoring has the capability to provide more than distance measurements, including vibration detection. Vibration can be associated with loose mechanical mounting of the sensor or possible mechanical issues with the machine that the sensor is mounted. That same laser distance sensor can also provide you with inclination angle measurement to help with the installation of the sensor or help detect when there’s a problem, such as when someone or something bumps the sensor out of alignment. What about ambient data, such as humidity? This could help detect or monitor for moisture ingress. Ambient pressure? It can be used to monitor the performance of fans or the condition of the filter elements on electrical enclosures.

Having access to condition monitoring data can help OEMs improve sensing capabilities of their machines, differentiating themselves from their competition. It can also help end users by providing them with real time monitoring of their assets; improving overall equipment efficiency and better predicting  and, thereby, eliminating unscheduled and costly machine downtime. These are just a few examples of the possibilities, and as market needs change, manufacturers of these devices can adapt to the market needs with new and improved functions, all thanks to smart device architecture.

Integrating smart devices to your control architecture

The most robust, cost effective, and reliable way of collecting this data is via the IO-Link communication protocol; the first internationally accepted open, vendor neutral, industrial bi-directional communications protocol that complies with IEC61131-9 standards. From there, this information can be directly passed to your machine controller, such as PLC, via fieldbus communication protocols, such as EtherNET/Ip, ProfiNET or EtherCAT, and to your SCADA / GUI applications via OPC/UA or JSON. There are also instances where wireless communications are used for special applications where devices are placed in hard to reach places using Bluetooth or WLAN.

In the fast paced ever changing world of industrial automation, condition monitoring data collection is increasingly more important. This data can be used in predictive maintenance measures to prevent costly and unscheduled downtime by monitoring vibration, inclination, and ambient data to help you stay ahead of the game.

Be Driven by Data and Decrease Downtime

Being “driven by data” is simply the act of making decisions based on real data instead of guessing or basing them on theoretical outcomes. Why one should do that, especially in manufacturing operations, is obvious. How it is done is not always so clear.

Here is how you can use a sensor, indicator light, and RFID to provide feedback that drives overall quality and efficiency.

 

Machine Condition Monitoring

You’ve heard the saying, “if it ain’t broke, don’t fix it.” However, broken machines cause downtime. What if there was a way to know when a machine is getting ready to fail, and you could fix it before it caused downtime? You can do that now!

The two main types of data measured in manufacturing applications are temperature and vibration. A sudden or gradual increase in either of these is typically an indicator that something is going wrong. Just having access to that data won’t stop the machine from failing, though. Combined with an indicator light and RFID, the sensor can provide real-time feedback to the operator, and the event can be documented on the RFID tag. The machine can then be adjusted or repaired during a planned maintenance period.

Managing Quality – A machine on its way to failure can produce parts that don’t meet quality standards. Fixing the problem before it affects production prevents scrap and rework and ensures the customer is getting a product with the quality they expect.

Managing Efficiency– Unplanned downtime costs thousands of dollars per minute in some industries. The time and resources required to deal with a failed machine far exceed the cost of the entire system designed to produce an early warning, provide indication, and document the event.

Quality and efficiency are the difference makers in manufacturing. That is, whoever makes the highest quality products most efficiently usually has the most profitable and sustainable business. Again, why is obvious, but how is the challenge. Hopefully, you can use the above data to make higher quality products more efficiently.

 

More to come! Here are the data-driven topics I will cover in my next blogs:

  • Part inspection and data collection for work in process
  • Using data to manage molds, dies, and machine tools

Adding Smart Condition Monitoring Sensors to Your PLC Control Systems Delivers Data in Real Time

Condition monitoring of critical components on machines delivers enormous benefits to productivity in a plant.  Rather than have a motor, pump, or compressor unexpectedly fail and the machine be inoperable until a replacement part is installed, condition monitoring of those critical pieces on the machine can provide warning signs that something is about to go terribly wrong. Vibration measurements on rotating equipment can detect when there is imbalance or degrade on rolling bearing elements. Temperature measurements can detect when a component is getting overheated and should be cooled down. Other environmental detections such as humidity and ambient pressure can alert someone to investigate why humidity or pressure is building up on a component or in an area. These measurement points are normally taken by specific accelerometers, temperature probes, humidity and pressure sensors and then analyzed through high end instruments with special analysis software. Typically, these instruments and software are separate from the PLC controls system. This means that even when the data indicates a future potential issue, steps need to be taken separately to stop the machine from running.

Using smart condition monitoring sensors with IO-Link allows these measured variables and alarms to be available directly onto the PLC system in real time. Some condition monitoring sensors now even have microprocessors onboard that immediately analyze the measured variables. The sensor can be configured for the measurement limit thresholds of the device it’s monitoring so that the sensor can issue a warning or alarm through the IO-Link communications channel to the PLC once those thresholds have been hit. That way, when a warning condition presents itself, the PLC can react immediately to it, whether that means sending an alert on a HMI, or stopping the machine from running altogether until the alarmed component is fixed or replaced.

Having the condition monitoring sensor on IO-Link has many advantages. As an IEC61131-9 standard, IO-Link is an open standard and not proprietary to any manufacturer. The protocol itself is on the sensor/actuator level and fieldbus independent. IO-Link allows the condition monitoring sensor to connect to Ethernet/IP, Profinet & Profibus, CC-Link & CC-Link IE Field, EtherCAT and TCP/IP networks regardless of PLC. Using an IO-Link master gateway, multiple smart condition monitoring sensors and other IO-Link devices can be connected to the controls network as a single node.

The picture above shows two condition monitoring sensors connected to a single address on the fieldbus network. In this example, a single gateway allows up to eight IO-Link condition monitoring sensors to be connected.

Through IO-Link, the PLC’s standard acyclic channel can be used to setup the parameters of the measured alarm conditions to match the specific device the sensor is monitoring. The PLC’s standard cyclic communications can then be used to monitor the alarm status bits from the condition monitoring sensor.  When an alarm threshold gets hit, the alarm status bit goes high and the PLC can then react in real time to control the machine. This relieves the burden of analyzing the sensor’s condition monitoring data from the PLC as the sensor is doing the work.