Talk to Me, One Machine Said to the Other

The promise is more than a marketing boast. Aided by microchip transmitters, heat sensors and a fast-growing form of wireless communication, the boast is a measurable fact.

Inside each Ocado delivery van is a SIM-card module the size of a postage stamp that monitors the air temperature. The sensor sends data to a computer used by fleet managers back at headquarters near London every few minutes.

Ocado says incidents of spoilage of goods have declined since the transmitters were installed last year.

“It has saved us time and given us more confidence in our real-time monitoring, as well as being a safety check for the driver,” said Paul Clarke, Ocado’s director of technology, who oversees a 300-person department that develops software and hardware for the retailer.

The drone of low-density conversation between Ocado’s trucks and headquarters in Hatfield is one example of machine-to-machine communication, a stream of consciousness based on semiconductors that is poised to reinvigorate the mobile industry.

Berg Insight, a research firm in Goteborg, Sweden, says the number of machine-to-machine devices using the world’s wireless networks reached 108 million in 2011 and will at least triple that by 2017. Ericsson, the leading maker of wireless network equipment, sees as many as 50 billion machines connected by 2020. Only 10 billion or so are likely to be cellphones and tablet computers. The rest will be machines, talking not to us, but to each other.

The combined level of robotic chatter on the world’s wireless networks — measured in the digital data load they exert on networks — is likely soon to exceed that generated by the sum of all human voice conversations taking place on wireless grids.

“I would say that is definitely possible within 10 years,” said Miguel Blockstrand, the director of Ericsson’s machine-to-machine division in Stockholm. “This is a ‘What if?’ kind of technology. People start to consider the potential, and the possibilities are endless.”

Machine-to-machine communications has been around for more than two decades, initially run on landline connections and used for controlling industrial processes remotely. With advances in mobile broadband speeds and smartphone computing, the same robotic conversations are now rapidly shifting to wireless networks.

When the total amount of data traffic generated by machines overtakes that created by human voice conversations — or possibly before — mobile operators will have to choose who waits in line to make a call or receive an e-mail — the machine or the human.

“It really does raise some quandaries for the operators,” said Tobias Ryberg, an analyst at Berg Insight. “Most mobile networks are set up for human communication, not for machines. So there will have to be a whole revamping of the system to make this possible.”

Currently, about a third of all machine-to-machine communication involves so-called smart utility meters, which perform duties like sending data on household electric and gas consumption to utilities; the utilities use the information to tailor production to actual demand. In Europe, all households in Sweden and Italy are equipped with smart meters, many of them running wirelessly. In Austria, a law will require five million homes to be equipped with smart electric meters by 2019.

Another third is taking place in the auto industry, through car and truck fleet management systems, which allow transport companies or corporate car managers to track their vehicles in real time, or that are used by emergency accident, repair and location services like General Motors’ OnStar system, now installed on a quarter of new GM vehicles. In Europe, similar technology is beginning to appear in preparation for 2015, when the eCall initiative, an E.U. law requiring all new cars to be equipped with wireless transmitters will take effect. The transmitters would automatically report accident data, as well as airbag deployment and location, to emergency responders