The second improvement would be to vibrate the payload less than other rockets do, easing design satellite requirements or improving reliability. Space launchers shake their payloads severely due to engine noise, aerodynamic effects and the impulse from the separation mechanisms. The H-IIA does that at least as much as its competitors, says Asada, adding that a new family would improve in all three areas.

Building that family will require funding, but Mitsubishi Heavy says the government should move while it has the advantage of a highly experienced workforce. About a quarter of Japan's space engineers—those who have had the most experience developing the H-II series and its predecessors—will retire by 2020, Asada said at the Japan International Aerospace Exhibition here last week. By 2020, an engineer who was 25 in the mid-1980s, during early work on the original H-II rocket, will be around 60.

So 2020 should be the target date for the entry into service of the H-X, Mitsubishi Heavy has told the government. JAXA has also designated 2020 as the target for flying the first H-X. Since development would take 6-8 years, there is little scope for delay.

Mitsubishi Heavy has handled H-IIA launches since 2007; JAXA conducted them before then.

Putting production and operation in the hands of one company would theoretically save money, and Mitsubishi Heavy did indeed simplify and improve the process, says Asada. Since the company was using the rockets itself, it could reduce the number of inspections. Now H-IIAs are checked just once. In another rationalization, design and development engineers follow each rocket from production to launch—it is not necessary to have completely separate teams for each phase.

Costs have not fallen, however, says Asada. In fact, they are a little higher than before, because each launch has resulted in demands to refine the H-IIA design. But costs would have been even higher had the government not given the launch work to Mitsubishi Heavy, he says. The government is evidently satisfied with the company's performance, because it has now decided to contract it to launch the heavier H-IIB rocket. Asada does not expect expenses to fall, though, and he declines to say why the government made the choice.

The H-IIA's payload is limited by the location of Tanegashima, which demands a greater impulse for orbital speeds (since the radial velocity of the Earth there is lower than at the equator) and because a high launch latitude results in an orbit with an equally high angle from the equator. In its standard H2A202 version, with two boosters, the H-IIA can lift 4 tons to this skewed geostationary transfer orbit. The satellite, on reaching the farthest point, must then burn an unusually large amount of fuel to reach the properly aligned geostationary track, adjusting its velocity by 1,830 meters per sec. (6,000 fps.). Customers are reluctant to accept that demand.

So Mitsubishi Heavy is working on an adaptation of the H-IIA, with a second stage that would remain with the satellite until it reached that distant point, coasting all the way, and would then fire again to correct the orbital inclination. The satellite would then need only to change its velocity by a more typical 1,500 meters per second to reach geosynchronous orbit. The modified rocket could deliver a 2.9-ton satellite in that manner but, for it to do so, the designers must minimize evaporation of the hydrogen fuel of the second stage during the 5-hr. coast. Asada says that applying a reflective white coating to the fuel tank, to keep the hydrogen cold, will help. The modified rocket will also need more electrical battery capacity and its thrust would have to be throttled.

H-IIAs and H-IIBs are used almost exclusively for Japanese government payloads, but in May the H-IIA performed the first mission for which it competed in the open market, launching a South Korean government satellite.