Slow Progress in Discrete Industries for Wireless Sensors

The Current Value Proposition for Wireless Sensors in Discrete Manufacturing Is Rooted in New Opportunities to Collect Manufacturing Data Where It Previously Was Not Possible

Share Print Related RSS

Phil BurgertBy Phil Burgert

As wireless sensors make inroads in discrete manufacturing, issues of signal latency and interference in some manufacturing environments continue to slow that process.

Among the advantages wireless sensors offer is the possibility of easier deployment in existing facilities and structures, reducing the cost to run cabling and create other infrastructures.

“Just eliminating the cost to plumb into existing or retrofit applications is the big motivator,” says Mark Spiering, application engineer for the energy and automation division of LEM. LEM, which in the past made only products that were hardwired, recently introduced its first Wi-Fi product. “We are relatively new to the technology,” says Spiering. “Most of our other products are integrated into our customers’ devices at the board level or in motor drives.”

Applications that use the wireless product include a customer facility that operates 20-ton presses and roll-forming machinery to manufacture 2 million truck wheels per year. The plant uses the LEM product to monitor individual presses for machine wear and tear based on increases in energy consumption and for scheduling maintenance.

Signal latency in energy monitoring is not an issue since the data broadcasts at 5-to-12-minute intervals, but Spiering says there have been issues with interference caused by metal in the factory. “The shop floor is relatively open, but there is so much metal that we have had some issues of dropouts that we’ve been able to address by being able to deploy extra nodes—repeater nodes to tie the network together,” he says.

Jim Bozas, wireless product manager for Adalet Wireless and Meriam Process Technologies says a key driver in the use of wireless sensors and networks is the cost of media involved in installing wired systems.

“There is a clear advantage in using wireless networks to gather data and monitor processes for industrial solutions,” he says. “With the rising costs of materials such as twisted-pair wire, fiberoptic cable and metal conduit, industrial wireless systems are a very efficient and reliable alternative to any conventional hardwired data network.”

Wireless can be an effective alternative to monitor the status of a switch or a proximity sensor although a significant disadvantage for certain factory automation applications is the average 500 msec to a full second required for a wireless device to communicate its status to a wireless receiver or a controller, says Bozas. This can be an issue for any sensor that uses a high switching-frequency application. The time latency for the transmitter in the field to reach its controller is not going to provide real-time status.

In some cases a complete packet may not reach the receiver during a transmission due to interference, and the transmitter must continue to send the information until the receiver has the complete packet.

This, Bozas reminds us, is a signal that is being broadcast by a license-free, low-power radio to avoid needing to purchase a license from government agencies, noting that 1 W is used for 900 MHz and 500 mW for 2.4 GHz.

“The main problem is speed is mandatory in many applications,” says Bozas. “Industrial wireless is most effectively used when monitoring sensors with a 4–20 mA output. Most devices using 4–20 usually change on a slow and gradual basis.”

Wireless is not yet the best solution for high-speed discrete applications, cautions Bozas, adding that he wouldn’t recommend a wireless solution for any application with inputs and outputs that change status in less than 1 sec.

Bob Gardner, wireless products engineer for Banner Engineering describes the current value proposition for wireless sensors in discrete manufacturing as rooted in new opportunities to collect manufacturing data where it previously was not possible. “This data could have been on a rotating machine where slip rings created maintenance issues or was simply too expensive to reach with wired infrastructure,” he says.

A valuable aspect of wireless is the mobility of the network to shift as requirements change, adds Gardner. “As budgets become increasingly tight, it will be an imperative that new equipment can be repurposed as application requirements change,” he says.

Mobility also enables testing of applications where an integrator can run a proof of a concept wireless network quickly and easily before the customer or end user commits to a wired installation.

On issues of technology and standards development suitable for discrete manufacturing at the sensor/actuator level, Gardner notes that end users and manufacturers want all wireless solutions to plug-and-play with each other.

“However, there are many reasons not to standardize all layers of wireless RF communication,” he argues. For example, he notes, in many applications 900 MHz is a more robust frequency for indoor plant-wide radio transmission. “The 2.4 GHz frequency band is crowded with handheld radios, Wi-Fi and even microwaves,” says Gardner. “There are pros and cons to both bands, but standards-based industrial wireless will be limited to 2.4 GHz, because it is recognized worldwide.”

Proprietary radios also can be tuned or optimized for a specific environment, and this flexibility wouldn’t be possible with a completely standardized radio specification, notes Gardner. “This being said, a standards-based gateway or network controller that includes certain protocols and specified techniques for interfacing to host devices makes perfect sense and doesn't jeopardize the wireless data stream,” he says.

Gardner concedes that data carried over a wire likely always will be faster than through the air. “However, there are many key applications where high-speed data collection at the sensing point or node is then being stored and forwarded to the network control device or gateway for processing,” he says. “Although it might take the gateway another 100 msec to have the identical count as the node, this does not affect the operational speed.”

Philip Burgert is a freelance writer, specializing in the technical trade media.

Share Print Reprints Permissions

What are your comments?

You cannot post comments until you have logged in. Login Here.

Comments

No one has commented on this page yet.

RSS feed for comments on this page | RSS feed for all comments